Automated polyp segmentation in colonoscopy videos is an essential computer-aided technology for early detection and removal of polyps. However, most existing video polyp segmentation methods are designed with pixel-level temporal learning mechanisms, at the cost of time-consuming frame-wise annotations. In this paper, we present VPSentry, a novel semi-supervised segmentation model with a sentry mechanism. Our model integrates a prototype memory to store the long-term spatiotemporal cues of colonoscopy videos. Moreover, we devise adaptive prototypes to capture and generalize critical representations from individual frames, enabling long-term temporal fusion across labeled and unlabeled frames. In addition, we propose a correlation dynamic propagation module that propagates information from prototypes to features while simultaneously extracting dynamic features to perceive variations in polyp details between adjacent frames. Since colonoscopy scenes may change among consecutive frames, we further employ a sentry mechanism to assess the inter-frame continuity. This mechanism guides the prototype memory updating and the correlation dynamic propagation, further facilitating robust temporal propagation and dynamic detail perception for semi-supervised learning of long-term colonoscopy video sequences. Extensive experiments on the large-scale SUN-SEG dataset demonstrate that our model achieves optimal segmentation performance with real-time inference efficiency.

Reliable zero-shot detection of out-of-distribution (OOD) inputs is critical for deploying vision-language models in open-world settings. However, the lack of labeled negatives in zero-shot OOD detection necessitates proxy signals that remain effective under distribution shift. Existing negative-label methods rely on a fixed set of textual proxies, which (i) sparsely sample the semantic space beyond in-distribution (ID) classes and (ii) remain static while only visual features drift, leading to cross-modal misalignment and unstable predictions. In this paper, we propose CoEvo, a training- and annotation-free test-time framework that performs bidirectional, sample-conditioned adaptation of both textual and visual proxies. Specifically, CoEvo introduces a proxy-aligned co-evolution mechanism to maintain two evolving proxy caches, which dynamically mines contextual textual negatives guided by test images and iteratively refines visual proxies, progressively realigning cross-modal similarities and enlarging local OOD margins. Finally, we dynamically re-weight the contributions of dual-modal proxies to obtain a calibrated OOD score that is robust to distribution shift. Extensive experiments on standard benchmarks demonstrate that CoEvo achieves state-of-the-art performance, improving AUROC by 1.33% and reducing FPR95 by 45.98% on ImageNet-1K compared to strong negative-label baselines.

Recent advances in deep learning have led to significant improvements in nuclei segmentation from histological images, particularly when labels of all classes are available simultaneously during training. However, in clinical practice, real-world scenarios require a model to perform well in an incremental learning setting, where we anticipate the model to achieve satisfactory performance on previously unseen data while effectively mitigating catastrophic forgetting of old classes. Most previous methods alleviate forgetting by distilling old class knowledge through prototypes; however, they fail to adequately capture fine-grained details to address the challenge of high class similarity, which is particularly severe in histological images. To overcome these limitations, we propose a novel incremental learning method for nuclei segmentation (we call it CiNuSeg), which is composed of two key innovative modules. First, we propose a new Anchor-driven Consistency Learning (ACL) module to construct multi-level class anchors within each sample to effectively capture fine structural and textural details of nuclei, thereby significantly mitigating forgetting. Second, we develop a Dual Region Regularization (DRR) module to suppress new class representations within old class regions while enhancing new class representations within new class regions, strengthening the model's ability to discriminate between different nuclei types and improving inter-class separability. We further introduce an Adaptive Temperature Tuning (ATT) strategy to dynamically balance model stability and plasticity. Extensive experiments conducted on benchmarking MoNuSAC and CoNSeP pathological datasets demonstrate the effectiveness of our method, consistently achieving better performance than SOTAs in different settings. Codes will be available upon publication.

Low-count positron emission tomography (PET) reconstruction is a challenging inverse problem due to severe degradations arising from Poisson noise, photon scarcity, and attenuation correction errors. Existing deep learning methods typically address these in the spatial domain with an undifferentiated optimization objective, making it difficult to disentangle overlapping artifacts and limiting correction effectiveness. In this work, we perform a Fourier-domain analysis and reveal that these degradations are spectrally separable: Poisson noise and photon scarcity cause high-frequency phase perturbations, while attenuation errors suppress low-frequency amplitude components. Leveraging this insight, we propose FourierPET, a Fourier-based unrolled reconstruction framework grounded in the Alternating Direction Method of Multipliers. It consists of three tailored modules: a spectral consistency module that enforces global frequency alignment to maintain data fidelity, an amplitude–phase correction module that decouples and compensates for high-frequency phase distortions and low-frequency amplitude suppression, and a dual adjustment module that accelerates convergence during iterative reconstruction. Extensive experiments demonstrate that FourierPET achieves state-of-the-art performance with significantly fewer parameters, while offering enhanced interpretability through frequency-aware correction.

Vision-Language Foundation Models (VLFMs) exhibit remarkable generalization, yet their direct application to medical ultrasound is severely hindered by a profound modality gap. The unique acoustic physics of ultrasound, characterized by speckle noise, shadowing, and heterogeneous textures, often degrades the performance of off-the-shelf VLFMs. To bridge this gap, we propose a novel Hybrid Tuning (HT) strategy for the parameter-efficient adaptation of CLIP-based models to ultrasound analysis. Instead of updating the pre-trained weights, HT freezes the visual backbone and integrates a specialized lightweight adapter. This adapter features a Frequency Filtering module to suppress domain-specific periodic artifacts and a Noise Estimation module to dynamically calibrate feature representations. Extensive evaluations across six multi-center datasets demonstrate that our HT-enhanced models significantly outperform existing state-of-the-art adapters and medical VLFMs in both segmentation and classification tasks. Notably, HT exhibits exceptional data efficiency in few-shot scenarios and robust cross-dataset generalization. Our findings prove that preserving pre-trained semantic priors while explicitly modeling ultrasound-specific noise is key to unlocking foundational intelligence in automated ultrasound diagnosis. The source code will be made publicly available.

• A new liver landmark detection dataset, L3D-2K, comprising 2,000 keyframes sourced from surgical videos with professional annotations. • A novel deep learning framework D2GPLand+ that utilizes RGB-D information for laparoscopic liver landmark detections. • Proposing the DPE module, which incorporates learnable prompts with contrastive learning to discriminate the geometric features of different landmark categories from depth clues. • Introducing the CUMamba block that concurrently conducts cross-modal interactions on spatial dimension and feature reparameterization on channel dimension for effective RGB-D fusion. • Introducing the AFA scheme to highlight anatomical structures by implicit and explicit edge emphasis and controlling detail levels. Laparoscopic liver surgery presents a highly intricate intraoperative environment with significant liver deformation, posing challenges for surgeons in locating critical liver structures. Anatomical liver landmarks can greatly assist surgeons in spatial perception in laparoscopic scenarios and facilitate preoperative-to-intraoperative registration. To advance research in liver landmark detection, we develop a new dataset called L3D-2K , comprising 2,000 keyframes with expert landmark annotations from surgical videos of 47 patients. Accordingly, we propose a baseline, D 2 GPLand+, which effectively leverages depth modality to boost landmark detection performance. Concretely, we introduce a Depth-aware Prompt Embedding (DPE) scheme, which dynamically extracts class-related global geometric cues with the guidance of self-supervised prompts from the SAM encoder. Further, a Cross-dimension Unified Mamba (CUMamba) block is designed to comprehensively incorporate RGB and depth features with the concurrent spatial and channel scanning mechanism. Besides, we bring out an Anatomical Feature Augmentation (AFA) module that captures anatomical cues and emphasizes key structures by optimizing feature granularity. For benchmarking purposes, we evaluate our method and 17 mainstream detection models on L3D, L3D-2K, and P2ILF datasets. Experimental results demonstrate that D 2 GPLand+ obtains superior performance on all three datasets. Our approach provides surgeons with guiding clues that facilitate surgical operations and decision-making in complex laparoscopic surgery. Our code and dataset are available at https://github.com/cuiruize/D2GPLand-Plus .

Pathological grading of laryngeal squamous cell carcinoma (LSCC) based on whole-slide image (WSI) is crucial for the diagnosis, treatment and prognosis. According to pathologists' knowledge, tumor regions are highly associated with grading. However, existing multiple instance learning (MIL) methods tend to overrepresent weakly relevant non-tumor regions and irrelevant background, leading to poor grading performance and interpretability. Motivated by the above problems, we propose an end-to-end knowledge-driven MIL network with hierarchical cluster-incorporated aware filtering, i.e. HCF-MIL. Firstly, we develop the tumor-guiding cluster filtering for feature representation, which awarely filters out irrelevant instance-level information and adaptively assigns learnable weights to tumor and non-tumor instances. Secondly, conventional mean-based and max-based aggregation primarily capture the overall patterns, neglecting the contributions of the most representative individual instances. Therefore, we propose a novel enhanced filtering aggregation learning strategy to strengthen hierarchical tumor-related feature representation. Through end-to-end optimization, HCF-MIL reduces model's entropy value and facilitates better alignment between decision-making process and diagnostic behaviors of pathologists. Experiments on larynx and multicentre datasets show that HCF-MIL significantly improves both pathological grading performance and interpretability, providing a strong foundation for reliable clinical deployment.

Real-time and realistic reconstruction of 3D dynamic surgical scenes from surgical videos is a novel and unique tool for surgical planning and intraoperative guidance. The 3D Gaussian splatting (GS), with its high rendering speed and reconstruction fidelity, has recently emerged as a promising technique for surgical scene reconstruction. However, existing GS-based methods still have two obvious shortcomings for realistic reconstruction. First, they largely struggle to capture localized yet intricate soft tissue deformations caused by complex instrument-tissue interactions. Second, they fail to model spatiotemporal coupling among Gaussian primitives for global adjustments during rapid perspective transformations, resulting in unstable reconstruction outputs. In this paper, we propose EndoRD-GS, an innovative approach that overcomes these two limitations through two core techniques: 1) periodic modulated Gaussian functions and 2) a new Biplane module. Specifically, our periodic modulated Gaussian functions incorporate meticulously designed modulations, significantly enhancing the representation of complex local tissue deformations. On the other hand, our Biplane module constructs spatiotemporal interactions among Gaussian primitives, enabling global adjustments and ensuring reliable scene reconstruction during rapid perspective transformations. Extensive experiments on three datasets demonstrate that our EndoRD-GS achieves superior performance in endoscopic scene reconstruction compared to state-of-the-art methods. The code is available at EndoRD-GS.

Cardiovascular health is vital to human well-being, and cardiac magnetic resonance (CMR) imaging is considered the clinical reference standard for diagnosing cardiovascular disease. However, its adoption is hindered by long scan times, complex contrasts, and inconsistent quality. While deep learning methods perform well on specific CMR imaging sequences, they often fail to generalize across modalities and sampling schemes. The lack of benchmarks for high-quality, fast CMR image reconstruction further limits technology comparison and adoption. The CMRxRecon2024 challenge, attracting over 200 teams from 18 countries, addressed these issues with two tasks: generalization to unseen modalities and robustness to diverse undersampling patterns. We introduced the largest public multi-modality CMR raw dataset, an open benchmarking platform, and shared code. Analysis of the best-performing solutions revealed that prompt-based adaptation and enhanced physics-driven consistency enabled strong cross-scenario performance. These findings establish principles for generalizable reconstruction models and advance clinically translatable AI in cardiovascular imaging.

Retinal image segmentation is essential for analyzing retinal structures like vessels and diagnosing retinopathy. However, the inherent intricacy of the retina, along with annotation scarcity and data heterogeneity, presents prevalent challenges in creating accurate and generalizable deep learning models. Diffusion models, while initially developed for image generation, have recently shown great promise for visual perception by leveraging the learned internal representations. However, these diffusion representations, which spread across network blocks (space) and diffusion timesteps (time), potentially suffer from issues like stochastic semantic distortion and cumulative structural blurring, compromising their semantic fidelity to the source image. In this paper, by delving into the generalization property of diffusion models, we propose a novel anchoring inversion strategy to derive diffusion representations that are semantically faithful to the source image from the deterministic trajectory. Furthermore, we introduce a time-space frequency-aware aggregation interpreter (T&S-FreqAgg) to aggregate the multi-scale and multi-timestep diffusion representations in a frequency-aware way for Domain Generalizable Semantic Segmentation (DGSS). Extensive experiments on nine public retinal image datasets demonstrate the superiority of our proposed framework, DiffDGSSv2, over state-of-the-art methods. Our code will be available at: https://github.com/Xyporz/DiffDGSSv2.

Automated lesion segmentation through breast ultrasound (BUS) images is an essential prerequisite in computer-aided diagnosis. However, the task of breast segmentation remains challenging, due to the time-consuming and labor-intensive process of acquiring precise labeled data, as well as severely ambiguous lesion boundaries and low contrast in BUS images. In this article, we propose a novel semi-supervised breast segmentation framework based on confidence-ranked features and bi-level prototypes (CoBiNet) to alleviate these issues. Our outputs are derived from two branches: classifier and projector. In the projector branch, we first rank the features by multilevel sampling to obtain multiple feature sets with different confidence levels. Then, these sets are progressed in two directions. One is to acquire local prototypes at each level by local sampling and perform trans-confidence level (TCL) contrastive learning. This encourages the low-confidence features to converge to the high-confidence features, which enhances the model's ability to recognize ambiguous regions. The other process is to generate more representative global prototypes by global sampling, followed by generating more reliable predictions and performing cross-guidance (CG) consistency learning with the classifier output predictions, facilitating knowledge transfer between the structure-aware projector and the category-discriminative classifier branches. Extensive experiments on two well-known public datasets, BUSI and UDIAT, demonstrate the superiority of our method over state-of-the-art approaches. Codes will be released upon publication.

Chinese poetry, a pinnacle of cultural expression, encapsulates human emotions and societal narratives in succinct, evocative language. Its unique blend of linguistic constraints and musicality makes analyzing its pronunciation's emotional features crucial for enriching children's linguistic prowess and artistic appreciation. To this end, we propose CPFformer, a novel deep learning framework, merges phonetics, sentiment analysis in order to analyze and predict the emotional feature of Chinese poetry pronunciation effectively. CPFformer, which consists of anomaly detection in spatial network (ADSN), spatial-temporal learning (STL), dimension segmentation and embedding (DSW), extraction of temporal attention (ETA), and encoder-decoder module (EDSM) modules, employs graph structures to capture the global and local consistency of emotional features across spatial and temporal, and a multiscale Mel feature extraction technique ensures comprehensive analysis of speech dynamics, enhancing emotional expression understanding. The mean square error (mse), mean absolute error (MAE), residual standard error (RSE), and $R$ -square ( $R^{2}$ ) of experiments reach 0.3572, 0.2486, 0.2014, and 0.9822, respectively, demonstrating its feasibility and effectiveness, exhibiting its superiority to the state-of-the-art approaches. The creation of a dedicated Chinese poetry pronunciation dataset marks a significant contribution, facilitating further research. The potential of CPFformer in speech technology and education heralds a new era, fostering the integration of traditional culture and artificial intelligence, and promoting the advancement of emotional literacy and smart learning environments. Its interdisciplinary implications promise exciting avenues for research and application.

With large-scale language models demonstrating superior capabilities in a wide range of downstream natural language processing tasks, the future trajectory of research in the field of text categorization faces increasing uncertainty. In this evolving paradigm of open-ended language modeling, where task delimitations are increasingly blurred, a pressing question arises: to what extent has text classification advanced under the full potential of large language model (LLM)? To address this pivotal inquiry, we introduce recurrent generative pre-trained transformer (RGPT), an adaptive boosting framework meticulously designed to craft a dedicated LLM for text classification. RGPT constructs a sequence of base learners by dynamically modulating the training data distribution and iteratively fine-tuning LLMs. These base learners are then progressively integrated, leveraging historical prediction trajectories to form a highly specialized text classification model. Extensive empirical evaluations demonstrate that RGPT surpasses eight state-of-the-art pretrained language models and seven cutting-edge LLMs across four benchmark datasets, achieving an average performance gain of 2.90%.

Singing-driven 3D head animation is a compelling yet underexplored task with broad applications in virtual avatars, entertainment, and education. Existing speech-driven approaches, which typically map audio directly to motion through implicit phoneme-to-viseme correspondences, often yield over-smoothed, emotionally flat, and semantically inconsistent results. These limitations render them inadequate for the unique demands of singing-driven animation. To address this challenge, we propose Think2Sing, a unified diffusion-based framework that integrates pretrained large language models to generate semantically consistent and temporally coherent 3D head animations conditioned on both lyrics and acoustics. Central to our framework is the introduction of motion subtitles, a structured, time-aligned representation generated via a Singing Chain-of-Thought process with acoustic-guided retrieval. These subtitles provide region-specific expressive cues that serve as interpretable priors for animation synthesis. We further formulate head animation as motion intensity prediction over key facial regions, enabling fine-grained control and more faithful expressive modeling. To support this paradigm, we construct the first multimodal singing dataset with synchronized 3D motion, acoustic descriptors, and aligned motion subtitles, enabling semantically grounded and expressive motion learning. Extensive experiments demonstrate that Think2Sing significantly outperforms state-of-the-art methods in realism, expressiveness, and emotional fidelity. Furthermore, our framework supports flexible subtitle-conditioned editing, enabling precise and user-controllable animation synthesis.

Tongue diagnosis is a vital tool in both Western and Traditional Chinese Medicine, providing key insights into a patient's health by analyzing tongue attributes. The COVID-19 pandemic has heightened the need for accurate remote medical assessments, emphasizing the importance of precise tongue attribute recognition via telehealth. To address this, we propose a Sign-Oriented multi-label Attributes Detection Framework. Our approach begins with an adaptive tongue feature extraction module that standardizes tongue images and mitigates environmental factors. This is followed by a Sign-oriented Network (SignNet) that identifies specific tongue attributes, emulating the diagnostic process of experienced practitioners and enabling comprehensive health evaluations. To validate our methodology, we developed an extensive tongue image dataset specifically designed for telemedicine. Unlike existing datasets, ours is tailored for remote diagnosis, with a comprehensive set of attribute labels. This dataset will be openly available, providing a valuable resource for research. Initial tests have shown improved accuracy in detecting various tongue attributes, highlighting our framework's potential as an essential tool for remote medical assessments.

In clinical imaging, medical segmentation networks typically require continually adapting to new data from multiple sites over time, as aggregating all data for learning at once can be impractical due to storage limitations and privacy concerns. However, existing methods basically overlook domain-specific characteristics and fall short of adequately capturing domain-invariant knowledge during continual learning, leading to undesired catastrophic forgetting of previous sites and inferior generalization to new sites. To tackle this issue, this paper introduces FR2Seg, to sufficiently exploit both domain-specific and domain-invariant knowledge for efficient continual learning with the aid of low-frequency cues. For the former aspect, we propose a Fourier style replay module to synthesize pseudo images with old-site styles for data augmentation during new-site training, effectively preventing catastrophic forgetting without sacrificing data privacy. For the latter, we present a Fourier adaptive consistency regularization to identify and constrain the optimization of domain-invariant parameters with explicit awareness of knowledge transferability across sites, ensuring excellent generalizability to new sites. Experimental results on two public datasets confirm our method's superiority over existing state-of-the-art continual learning methods.

Elaborating a series of intermediate reasoning steps significantly improves the ability of large language models (LLMs) to solve complex problems, as such steps would evoke LLMs to think sequentially. However, human sarcasm understanding is often considered an intuitive and holistic cognitive process, in which various linguistic, contextual, and emotional cues are integrated to form a comprehensive understanding, in a way that does not necessarily follow a step-by-step fashion. To verify the validity of this argument, we introduce a new prompting framework (called SarcasmCue) containing four sub-methods, viz. chain of contradiction (CoC), graph of cues (GoC), bagging of cues (BoC) and tensor of cues (ToC), which elicits LLMs to detect human sarcasm by considering sequential and non-sequential prompting methods. Through a comprehensive empirical comparison on four benchmarks, we highlight three key findings: (1) CoC and GoC show superior performance with more advanced models like GPT-4 and Claude 3.5, with an improvement of 3.5%. (2) ToC significantly outperforms other methods when smaller LLMs are evaluated, boosting the F1 score by 29.7% over the best baseline. (3) Our proposed framework consistently pushes the state-of-the-art (i.e., ToT) by 4.2%, 2.0%, 29.7%, and 58.2% in F1 scores across four datasets. This demonstrates the effectiveness and stability of the proposed framework.

Accurate automatic breast ultrasound (BUS) image segmentation is essential for early breast cancer screening and diagnosis. However, it remains challenging owing to (1) breast lesions of various scale and shape, (2) ambiguous boundaries caused by speckle noise and artifacts, and (3) the scarcity of high-quality annotations. Most existing semi-supervised methods employ the mean-teacher architecture, which merely learns semantic information within a single image and heavily relies on the performance of the teacher model. Therefore, we present a novel cross-image semantic correlation semi-supervised framework, named CSC-PA, to improve the performance of BUS image segmentation. CSC-PA is trained based on a single network, which integrates a foreground prototype attention (FPA) and an edge prototype attention (EPA). Specifically, FPA transfers complementary foreground information for more stable and complete lesion segmentation. On the other hand, EPA enhances edge features of lesions by using edge prototype, where an adaptive edge container is proposed to store global edge features and generate the edge prototype. Additionally, we introduce a pixel affinity loss (PAL) to exploit previously ignored contextual correlation in supervision, which further improves performance on edges. Extensive experiments on two benchmark BUS datasets demonstrate that our model outperforms other state-of-the-art methods under different partition protocols. Codes are available at https://github.com/shdkdh/CSC-PA.

The self-attention mechanism of Transformer has gained considerable attention for its potential in modeling long-term temporal dependencies in electroencephalogram (EEG) signals. Despite recent advancements, Transformer-based decoding methods often neglect the explicit temporal priors inherent in EEG signals, i.e., the dependency between tokens tends to diminish as their relative temporal distance increases, which limits the efficacy of these decoding approaches. Inspired by the recent Retentive Network (RetNet), we develop a novel convolutional retentive network for EEG decoding (RetEEG), which integrates temporal priors into the self-attention mechanism to tackle the above challenge. Specifically, RetEEG incorporates a convolution module to capture local EEG features and cascades multiple bidirectional retention modules to learn global correlations within these local features. The bidirectional retention module introduces a temporal decay matrix, which imparts prior knowledge to the self-attention mechanism by accounting for variations in relative temporal distances. To validate the superiority of RetEEG, we conducted experiments on two publicly available EEG datasets. The experimental results empirically demonstrate that the proposed RetEEG achieves the state-of-the-art EEG decoding performance. Code is available at: https://github.com/kfhss/RetEEG.

Accurate segmentation of cardiac chambers in echocardiography sequences is crucial for the quantitative analysis of cardiac function, aiding in clinical diagnosis and treatment. The imaging noise, artifacts, and the deformation and motion of the heart pose challenges to segmentation algorithms. While existing methods based on convolutional neural networks, Transformers, and space-time memory networks have improved segmentation accuracy, they often struggle with the trade-off between capturing long-range spatiotemporal dependencies and maintaining computational efficiency with fine-grained feature representation. In this paper, we introduce GDKVM, a novel architecture for echocardiography video segmentation. The model employs Linear Key-Value Association (LKVA) to effectively model inter-frame correlations, and introduces Gated Delta Rule (GDR) to efficiently store intermediate memory states. Key-Pixel Feature Fusion (KPFF) module is designed to integrate local and global features at multiple scales, enhancing robustness against boundary blurring and noise interference. We validated GDKVM on two mainstream echocardiography video datasets (CAMUS and EchoNet-Dynamic) and compared it with various state-of-the-art methods. Experimental results show that GDKVM outperforms existing approaches in terms of segmentation accuracy and robustness, while ensuring real-time performance. Code is available at https://github.com/wangrui2025/GDKVM.

Out-of-distribution (OOD) detection is crucial for ensuring the reliability and safety of machine learning models in real-world applications. While zero-shot OOD detection, which requires no training on in-distribution (ID) data, has become feasible with the emergence of vision-language models like CLIP, existing methods primarily focus on semantic matching and fail to fully capture distributional discrepancies. To address these limitations, we propose OT-DETECTOR, a novel framework that employs Optimal Transport (OT) to quantify both semantic and distributional discrepancies between test samples and ID labels. Specifically, we introduce cross-modal transport mass and transport cost as semantic-wise and distribution-wise OOD scores, respectively, enabling more robust detection of OOD samples. Additionally, we present a semantic-aware content refinement (SaCR) module, which utilizes semantic cues from ID labels to amplify the distributional discrepancy between ID and hard OOD samples. Extensive experiments on several benchmarks demonstrate that OT-DETECTOR achieves state-of-the-art performance across various OOD detection tasks, particularly in challenging hard-OOD scenarios.

Few-shot learning (FSL) addresses the challenge of classifying novel classes with limited training samples. While some methods leverage semantic knowledge from smaller-scale models to mitigate data scarcity, these approaches often introduce noise and bias due to the data’s inherent simplicity. In this paper, we propose a novel framework, Synergistic Knowledge Transfer (SynTrans), which effectively transfers diverse and complementary knowledge from large multimodal models to empower the off-the-shelf few-shot learner. Specifically, SynTrans employs CLIP as a robust teacher and uses a few-shot vision encoder as a weak student, distilling semantic-aligned visual knowledge via an unsupervised proxy task. Subsequently, a training-free synergistic knowledge mining module facilitates collaboration among large multimodal models to extract high-quality semantic knowledge. Building upon this, a visual-semantic bridging module enables bi-directional knowledge transfer between visual and semantic spaces, transforming explicit visual and implicit semantic knowledge into category-specific classifier weights. Finally, SynTrans introduces a visual weight generator and a semantic weight reconstructor to adaptively construct optimal multimodal FSL classifiers. Experimental results on four FSL datasets demonstrate that SynTrans, even when paired with a simple few-shot vision encoder, significantly outperforms current state-of-the-art methods.

Automatic echocardiography video segmentation is a powerful tool for improving the accuracy of cardiovascular function assessment. However, it remains a challenging task owing to (1) extensive speckle noise and blurred boundaries, (2) dramatic shape variations of targeting structures across frames, and (3) limited labeled data due to the high cost of annotation. In this paper, we present a novel semi-supervised segmentation model based on Vision Mamba (Vim) to comprehensively tackle these challenges; we call it EchoVim. Our framework introduces three technical innovations: First, a bidirectional inference mechanism (BIM) which can propagate label information bidirectionally from end-diastolic (ED) and end-systolic (ES) frames to generate pseudo-labels, coupled with confidence-aware dynamic updating to progressively refine supervision signals. Second, a dynamic interaction temporal alignment (DITA) module that establishes anatomical correspondence across frames by adaptively enhancing features near temporally stable regions while suppressing motion-irrelevant artifacts, effectively addressing variations in cardiac shape. Third, a semantic token-attentive refinement (STR) module that constructs low-rank semantic tokens to encode cardiac structure priors, utilizing attention-guided nonlinear transformations to disentangle speckle noise from true anatomical patterns. We conduct extensive experiments on two benchmarking echocardiography video datasets: CAMUS and EchoNet-Dynamic, and the results demonstrate that our method outperforms existing state-of-the-art approaches with real-time inference. Codes are available at https://github.com/guojx2255/EchoVim.

Automatic segmentation of echocardiography videos is crucial for computer-aided cardiovascular function assessment in clinical practice. However, it is a challenging task owing to the existence of massive speckle noise, the large shape variations of heart structures between frames, and limited annotations. In this paper, we propose a novel semi-supervised video segmentation model to comprehensively meet these challenges. The proposed approach has two key techniques. First, we propose a dual-stream architecture that processes spatial and temporal features through separate pathways to capture structural details and motion patterns, then enhances spatiotemporal representations by interacting these decomposed features with query features generated from the original input. Second, as speckle noise primarily concentrates in high-frequency regions, we extend the traditional dilated convolution from a frequency perspective, enabling it to adaptively adjust the dilation rate and convolution kernel weights based on high frequency speckle noise information. This enables the network to focus on specific frequency bands, thereby enhancing its ability to capture both low-frequency context and high-frequency local details. Extensive experiments on the CAMUS and EchoNet-Dynamic datasets demonstrate that our method outperforms existing state-of-the-art methods in terms of both accuracy and inference speed. Codes are available at https://github.com/guojx2255/HSCA-SDC.

With the continuous application of single-cell data, effectively correcting batch effects and accurately identifying cell types has emerged as a critical challenge in biomedical research. However, existing methods often struggle to disentangle technical effects from genuine biological variation, limiting their performance on heterogeneous datasets. Here, we introduce single-cell Batch Correction Network (scBCN), an integration framework that combines robust inter-batch similar cluster identification with a deep residual neural network to correct batch effects while preserving biological variability. To evaluate the performance of scBCN, we conduct benchmarking experiments on various simulated and real datasets, demonstrating its superiority in both batch correction and biological variation conservation. Furthermore, scBCN shows its applicability in cross-species and cross-omics data integration, underscoring its potential for uncovering and characterizing cell type-specific gene expression patterns.

Automatic assessment of driver autonomic nervous system conditions is crucial for enhancing driving safety and healthcare. We present a novel approach that combines a dual optical fiber sensor system and a sophisticated deep learning framework, VHDP, with a strong emphasis on graph learning and spatiotemporal modeling techniques. The proposed fiber interferometer based dual optical fiber sensor system can effectively monitor driver vital signs in various environments. The VHDP framework, a significant innovation in deep learning, first utilizes the EMGLCN module to extract spatial and temporal features from the acquired heart rate variability (HRV) data for graph modeling. Then, through the dynamic spatial-temporal multi-graph method and the temporal-awareness attention module (TAA), it captures cross-time and dimensional correlations. Finally, the prior knowledge guided recalibration fusion module (PKGRF) generates accurate outputs. Experimental results show that the mean square error (MSE), mean absolute error (MAE) and R-square (R2) reach 2.354, 0.896 and 0.9857 respectively, outperforming state-of-the-art approaches. This work not only provides a new method for long-term driver HRV assessment and forecasting but also demonstrates the potential of graph learning and spatiotemporal modeling in the fields of medical monitoring and artificial intelligence, offering valuable insights for the development of portable vital signs monitoring devices in the context of the Internet of Medical Things (IoMT).

In the era of large language models (LLMs), tasks associated with “System I” cognition—those that are fast, automatic, and intuitive, such as sentiment analysis and text classification—are often considered effectively solved. However, sarcasm remains a persistent challenge. As a subtle and complex linguistic phenomenon, sarcasm frequently involves rhetorical devices such as hyperbole and figurative language to express implicit sentiments and intentions, demanding a higher level of abstraction and pragmatic reasoning than standard sentiment analysis. This raises concerns about whether current claims of LLM success extend robustly to the domain of sarcasm understanding. To systematically investigate this issue, we introduce a new high-quality multi-modal sarcasm detection dataset, termedAMSD, and construct a comprehensive evaluation benchmark,SarcasmBench. Our benchmark encompasses 16 state-of-the-art (SOTA) LLMs and 8 strong pretrained language models (PLMs), evaluated across six widely-used textual sarcasm datasets and three multi-modal sarcasm benchmarks. We adopt three popular prompting paradigms: zero-shot input/output (IO) prompting, few-shot IO prompting, and chain-of-thought (CoT) prompting. Our extensive experiments yield three key findings: (1) current LLMs underperform supervised PLMs based sarcasm detection baselines. This suggests that significant efforts are still required to improve LLMs' understanding of human sarcasm. (2) GPT-4 and Gemini 2.0 consistently and significantly outperforms other LLMs across various prompting methods. (3) Few-shot IO prompting method outperforms the other two methods: zero-shot IO and few-shot CoT. We hope this benchmark will serve as a valuable resource for the research community and inspire future work toward more robust and human-aligned sarcasm understanding.

6D object pose estimation from a single RGB-D image is a fundamental problem in computer vision and robot manipulation. Despite recent advancements, existing methods still suffer several limitations. First of all, the object shape representation extracted from the depth map is often less expressive because the object point cloud parsed from the depth map is highly incomplete due to the object self-occlusion and noisy due to the sensor artifacts. This shape representation issue further intensifies when lacking sufficient labeled data for model training, which unfortunately is another typical problem for object pose estimation considering the heavy annotation cost for real-world pose labeling. In this study, we propose to tackle the above issues in a unified way. First, we enhance the object shape representation from the partial point cloud with a novel canonical shape reconstruction module, in which an implicit canonical frame is established by incorporating the SE(3) equivariance, achieving implicit feature alignment of the partial point cloud inputs, leading to robust shape recovery. Second, based on the enhanced object representation, we further utilize the de-canonicalized and pose-dependent completed object shape as the training signal, and develop a novel weakly-supervised learning framework to leverage both labeled synthetic data and unlabeled real data to train the pose estimation model in a label-efficient way. Extensive experiments on three widely used benchmarks demonstrate the effectiveness, and superiority of our framework over state-of-the-art methods.

Recently, functional magnetic resonance imaging (fMRI)-based brain networks have been shown to be an effective diagnostic tool with great potential for accurately detecting autism spectrum disorders (ASD). Meanwhile, the successful use of graph convolution networks (GCNs) methods based on fMRI information has improved the classification accuracy of ASD. However, many graph convolution-based methods do not fully utilize the topological information of the brain functional connectivity network (BFCN) or ignore the effect of non-imaging information. Therefore, we propose a hierarchical graph embedding model that leverage both the topological information of the BFCN and the non-imaging information of the subjects to improve the classification accuracy. Specifically, our model first use the Infomax Module to automatically identify embedded features in regions of interests (ROIs) in the brain. Then, these features, along with non-imaging information, is used to construct a population graph model. Finally, we design a graph convolution framework to propagate and aggregate the node features and obtain the results for ASD detection. Our model takes into account both the significance of the BFCN to individual subjects and relationships between subjects in the population graph. The model performed autism detection using the Autism Brain Imaging Data Exchange (ABIDE) dataset and obtained an average accuracy of 77.2% and an AUC of 87.2%. These results exceed those of the baseline approach. Through extensive experiments, we demonstrate the competitiveness, robustness and effectiveness of our model in aiding ASD diagnosis.

Diffusion models have garnered significant attention for MRI Super-Resolution (SR) and have achieved promising results. However, existing diffusion-based SR models face two formidable challenges: 1) insufficient exploitation of complementary information from multi-contrast images, which hinders the faithful reconstruction of texture details and anatomical structures; and 2) reliance on fixed magnification factors, such as 2× or 4×, which is impractical for clinical scenarios that require arbitrary scale magnification. To circumvent these issues, this paper introduces IM-Diff, an implicit multi-contrast diffusion model for arbitrary-scale MRI SR, leveraging the merits of both multi-contrast information and the continuous nature of implicit neural representation (INR). Firstly, we propose an innovative hierarchical multi-contrast fusion (HMF) module with reference-aware cross Mamba (RCM) to effectively incorporate target-relevant information from the reference image into the target image, while ensuring a substantial receptive field with computational efficiency. Secondly, we introduce multiple wavelet INR magnification (WINRM) modules into the denoising process by integrating the wavelet implicit neural non-linearity, enabling effective learning of continuous representations of MR images. The involved wavelet activation enhances space-frequency concentration, further bolstering representation accuracy and robustness in INR. Extensive experiments on three public datasets demonstrate the superiority of our method over existing state-of-the-art SR models across various magnification factors.

The precise segmentation of different brain regions and tissues is usually a prerequisite for the detection and diagnosis of various neurological disorders in neuroscience. Considering the abundance of functional and structural dual-modality information for positron emission tomography/magnetic resonance (PET/MR) images, we propose a novel 3D whole-brain segmentation network with a cross-fusion mechanism introduced to obtain 45 brain regions. Specifically, the network processes PET and MR images simultaneously, employing UX-Net and a cross-fusion block for feature extraction and fusion in the encoder. We test our method by comparing it with other deep learning-based methods, including 3DUXNET, SwinUNETR, UNETR, nnFormer, UNet3D, NestedUNet, ResUNet, and VNet. The experimental results demonstrate that the proposed method achieves better segmentation performance in terms of both visual and quantitative evaluation metrics and achieves more precise segmentation in three views while preserving fine details. In particular, the proposed method achieves superior quantitative results, with a Dice coefficient of 85.73% 0.01%, a Jaccard index of 76.68% 0.02%, a sensitivity of 85.00% 0.01%, a precision of 83.26% 0.03% and a Hausdorff distance (HD) of 4.4885 14.85%. Moreover, the distribution and correlation of the SUV in the volume of interest (VOI) are also evaluated (PCC > 0.9), indicating consistency with the ground truth and the superiority of the proposed method. In future work, we will utilize our whole-brain segmentation method in clinical practice to assist doctors in accurately diagnosing and treating brain diseases.

In multi-modal magnetic resonance imaging (MRI), the tasks of imputing or reconstructing the target modality share a common obstacle: the accurate modeling of fine-grained inter-modal differences, which has been sparingly addressed in current literature. These differences stem from two sources: 1) spatial misalignment remaining after coarse registration and 2) structural distinction arising from modality-specific signal manifestations. This paper integrates the previously separate research trajectories of cross-modality synthesis (CMS) and multi-contrast super-resolution (MCSR) to address this pervasive challenge within a unified framework. Connected through generalized down-sampling ratios, this unification not only emphasizes their common goal in reducing structural differences, but also identifies the key task distinguishing MCSR from CMS: modeling the structural distinctions using the limited information from the misaligned target input. Specifically, we propose a composite network architecture with several key components: a label correction module to align the coordinates of multi-modal training pairs, a CMS module serving as the base model, an SR branch to handle target inputs, and a difference projection discriminator for structural distinction-centered adversarial training. When training the SR branch as the generator, the adversarial learning is enhanced with distinction-aware incremental modulation to ensure better-controlled generation. Moreover, the SR branch integrates deformable convolutions to address cross-modal spatial misalignment at the feature level. Experiments conducted on three public datasets demonstrate that our approach effectively balances structural accuracy and realism, exhibiting overall superiority in comprehensive evaluations for both tasks over current state-of-the-art approaches. The code is available at https://github.com/papshare/FGDL.

In augmented reality (AR)-guided surgical navigation, preoperative organ models are superimposed onto the patient's intraoperative anatomy to visualize critical structures such as vessels and tumors. Accurate deformation modeling is essential to maintain the reliability of AR overlays by ensuring alignment between preoperative models and the dynamically changing anatomy. Although the finite element method (FEM) offers physically plausible modeling, its high computational cost limits intraoperative applicability. Moreover, existing algorithms often fail to handle large anatomical changes, such as those induced by pneumoperitoneum or ligament dissection, leading to inaccurate anatomical correspondences and compromised AR guidance. To address these challenges, we propose a data-driven biomechanics algorithm that preserves FEM-level accuracy while improving computational efficiency. In addition, we introduce a novel human-in-the-loop mechanism into the deformation modeling process. This enables surgeons to interactively provide prompts to correct anatomical misalignments, thereby incorporating clinical expertise and allowing the model to adapt dynamically to complex surgical scenarios. Experiments on a publicly available dataset demonstrate that our algorithm achieves a mean target registration error of 3.42 mm. Incorporating surgeon prompts through the interactive framework further reduces the error to 2.78 mm, surpassing state-of-the-art methods in volumetric accuracy. These results highlight the ability of our framework to deliver efficient and accurate deformation modeling while enhancing surgeon-algorithm collaboration, paving the way for safer and more reliable computer-assisted surgeries.

Medical vision-language alignment through cross-modal contrastive learning shows promising performance in image-text matching tasks, such as retrieval and zero-shot classification. However, conventional cross-modal contrastive learning (CLIP-based) methods suffer from suboptimal visual representation capabilities, which also limits their effectiveness in vision-language alignment. In contrast, although the models pretrained via multimodal masked modeling struggle with direct cross-modal matching, they excel in visual representation. To address this contradiction, we propose ALTA (ALign Through Adapting), an efficient medical vision-language alignment method that utilizes only about 8% of the trainable parameters and less than 1/5 of the computational consumption required for masked record modeling. ALTA achieves superior performance in vision-language matching tasks like retrieval and zero-shot classification by adapting the pretrained vision model from masked record modeling. Additionally, we integrate temporal-multiview radiograph inputs to enhance the information consistency between radiographs and their corresponding descriptions in reports, further improving the vision-language alignment. Experimental evaluations show that ALTA outperforms the best-performing counterpart by over 4% absolute points in text-to-image accuracy and approximately 6% absolute points in image-to-text retrieval accuracy. The adaptation of vision-language models during efficient alignment also promotes better vision and language understanding. Code is publicly available at https://github.com/DopamineLcy/ALTA.

Automatic and accurate classification of breast cancer in multimodal ultrasound images is crucial to improve patients' diagnosis and treatment effect and save medical resources. Methodologically, the fusion of multimodal ultrasound images often encounters challenges such as misalignment, limited utilization of complementary information, poor interpretability in feature fusion, and imbalances in sample categories. To solve these problems, we propose a feature alignment mutual attention fusion method (FAMF-Net), which consists of a region awareness alignment (RAA) block, a mutual attention fusion (MAF) block, and a reinforcement learning-based dynamic optimization strategy(RDO). Specifically, RAA achieves region awareness through class activation mapping and performs translation transformation to achieve feature alignment. When MAF utilizes a mutual attention mechanism for feature interaction fusion, it mines edge and color features separately in B-mode and shear wave elastography images, enhancing the complementarity of features and improving interpretability. Finally, RDO uses the distribution of samples and prediction probabilities during training as the state of reinforcement learning to dynamically optimize the weights of the loss function, thereby solving the problem of class imbalance. The experimental results based on our clinically obtained dataset demonstrate the effectiveness of the proposed method. Our code will be available at: https://github.com/Magnety/Multi_modal_Image.

Liver registration by overlaying preoperative 3D models onto intraoperative 2D frames can assist surgeons in perceiving the spatial anatomy of the liver clearly for a higher surgical success rate. Existing registration methods rely heavily on anatomical landmark-based workflows, which encounter two major limitations: 1) ambiguous landmark definitions fail to provide efficient markers for registration; 2) insufficient integration of intraoperative liver visual information in shape deformation modeling. To address these challenges, in this paper, we propose a landmark-free preoperative-to-intraoperative registration framework utilizing effective self-supervised learning, termed Self-P2IR. This framework transforms the conventional 3D-2D workflow into a 3D-3D registration pipeline, which is then decoupled into rigid and non-rigid registration subtasks. Self-P2IR first introduces a feature-disentangled transformer to learn robust correspondences for recovering rigid transformations. Further, a structure-regularized deformation network is designed to adjust the preoperative model to align with the intraoperative liver surface. This network captures structural correlations through geometry similarity modeling in a low-rank transformer network. To facilitate the validation of the registration performance, we also construct an in-vivo registration dataset containing liver resection videos of 21 patients, called P2I-LReg, which contains 346 keyframes that provide a global view of the liver together with liver mask annotations and calibrated camera intrinsic parameters. Extensive experiments and user studies on both synthetic and in-vivo datasets demonstrate the superiority and potential clinical applicability of our method. The code and dataset are available at https://github.com/junzastar/Self-P2IR.

The Facial Expression Recognition (FER) technique has increasingly matured over time. However, recognizing facial expressions in wild environments poses great challenges in achieving promising performance. The main obstacles arise from various factors, such as illumination changes, head pose variations, and occlusions. To overcome interferences from external environments and improve recognition accuracy, we propose a novel Quaternion Wavelet TRansformer (QWTR) model for FER in the wild. Specifically, we present a Quaternion Value Transformer (QVT) network that combines quaternion multi-head attention with quaternion CNN to capture emotional cues from global and local perception. To preserve the color structure while enhancing image contrast and brightness, we introduce a Quaternion Histogram Equalization (QHE) representation to transform color images into quaternion matrices representation. After that, to alleviate the impact of head pose and occlusion together with feature redundancy, a Quaternion Wavelet Feature Selection (QWFS) scheme is designed to decompose quaternion features and select the most correlated signals. Extensive experiments have been conducted on four in-the-wild FER datasets and several specific FER benchmarks under various conditions. The qualitative and quantitative results demonstrate thatQWTRoutperforms other state-of-the-art methods in FER benchmarks, e.g., 68.37% vs. 66.31% accuracy on the AffectNet dataset.

The widely deployed ways to capture a set of unorganized points, e.g., merged laser scans, fusion of depth images, and structure-from- , usually yield a 3-D noisy point cloud. Accurate normal estimation for the noisy point cloud makes a crucial contribution to the success of various applications. However, the existing normal estimation wisdoms strive to meet a conflicting goal of simultaneously performing normal filtering and preserving surface features, which inevitably leads to inaccurate estimation results. We propose a normal estimation neural network (Norest-Net), which regards normal filtering and feature preservation as two separate tasks, so that each one is specialized rather than traded off. For full noise removal, we present a normal filtering network (NF-Net) branch by learning from the noisy height map descriptor (HMD) of each point to the ground-truth (GT) point normal; for surface feature recovery, we construct a normal refinement network (NR-Net) branch by learning from the bilaterally defiltered point normal descriptor (B-DPND) to the GT point normal. Moreover, NR-Net is detachable to be incorporated into the existing normal estimation methods to boost their performances. Norest-Net shows clear improvements over the state of the arts in both feature preservation and noise robustness on synthetic and real-world captured point clouds.

Artifacts are prevalent in Optical Coherence Tomography Angiography (OCTA) images, which probably interfere doctor’s diagnosis and greatly limit its utility. Therefore, it is desirable to segment artifacts and assess quality when using them for diagnosis. In this article, we propose an end-to-end network (named CCM-Net: C ontrastive and C onsistent M ulti-task Network) to jointly address artifact segmentation and quality classification of OCTA images. We first devise multiple Task-Specific Attention Blocks to integrate deep features at different CNN layers for segmenting artifacts and classifying the quality of the input OCTA image. In this way, the weights of different deep features can be automatically learned and are not the same for the two tasks. Moreover, we devise a contrastive loss and a consistency loss to leverage sample relations for further enhancing prediction accuracy. Specifically, given an input OCTA image, we first augment it with a color jitter and select another OCTA image with the same quality classification label. We then design a contrastive loss so that the segmentation results of the input OCTA image are similar to its enhanced OCTA image, while the segmentation results of the two selected OCTA images are not similar. Besides, we devise a consistency loss on the classification results of the three images, because we can find that these images have the same quality classification labels. Experiments on an in-house OCTA dataset (Multi-OCTA) demonstrate that the proposed CCM-Net outperforms state-of-the-art methods.

Although pre-trained large-scale generative models StyleGAN series have proven to be effective in various editing and translation tasks, they are limited to pre-defined fixed aspect ratio. To overcome this limitation, we propose StyleGAN-$\infty$∞, a model that enables pre-trained StyleGAN to perform arbitrary-ratio conditional synthesis. Our key insight is to distill the expressive StyleGAN features into a StyleBook, such that an arbitrary-ratio condition can be translated to other forms by properly assembling pre-defined StyleBook vectors. To learn and leverage the StyleBook, we employ a network with three distinct stages, each corresponding to StyleBook extraction, StyleBook correspondence learning, and arbitrary-ratio synthesis. Extensive experiments on various conditional synthesis tasks, like super-resolution, sketch synthesis, and semantic synthesis, demonstrate superior performances over state-of-the-art image-to-image translation methods. Moreover, our model can easily generate megapixel images in diverse modalities by taking advantage of different pre-trained StyleGAN models.

Traditional human neural radiance fields often overlook crucial body semantics, resulting in ambiguous reconstructions, particularly in occluded regions. To address this problem, we propose the Super-Semantic Disentangled Neural Renderer (SSD-NeRF), which employs rich regional semantic priors to enhance human rendering accuracy. This approach initiates with a Visible-Invisible Semantic Propagation module, ensuring coherent semantic assignment to occluded parts based on visible body segments. Furthermore, a Region-Wise Texture Propagation module independently extends textures from visible to occluded areas within semantic regions, thereby avoiding irrelevant texture mixtures and preserving semantic consistency. Additionally, a view-aware curricular learning approach is integrated to bolster the model's robustness and output quality across different viewpoints. Extensive evaluations confirm that SSD-NeRF surpasses leading methods, particularly in generating quality and structurally semantic reconstructions of unseen or occluded views and poses.

Semi-supervised learning (SSL) is a powerful tool to address the challenge of insufficient annotated data in medical segmentation problems. However, existing semi-supervised methods mainly rely on internal knowledge for pseudo labeling, which is biased due to the distribution mismatch between the highly imbalanced labeled and unlabeled data. Segmenting left atrial appendage (LAA) from transesophageal echocardiogram (TEE) images is a typical medical image segmentation task featured by scarcity of professional annotations and diverse data distributions, for which existing SSL models cannot achieve satisfactory performance. In this paper, we propose a novel strategy to mitigate the inherent challenge of distribution mismatch in SSL by, for the first time, incorporating a large foundation model (i.e. SAM in our implementation) into an SSL model to improve the quality of pseudo labels. We further propose a new self-reconstruction mechanism to generate both noise-resilient prompts to demonically improve SAM’s generalization capability over TEE images and self-perturbations to stabilize the training process and reduce the impact of noisy labels. We conduct extensive experiments on an in-house TEE dataset; experimental results demonstrate that our method achieves better performance than state-of-the-art SSL models.

Automatic polyp segmentation from colonoscopy videos is a critical task for the development of computer-aided screening and diagnosis systems. However, accurate and real-time video polyp segmentation (VPS) is a very challenging task due to low contrast between background and polyps and frame-to-frame dramatic variations in colonoscopy videos. We propose a novel embedding-unleashing framework consisting of a proposal-generative network (PGN) and an appearance-embedding network (AEN) to comprehensively address these challenges. Our framework, for the first time, models VPS as an appearance-level semantic embedding process to facilitate generate more global information to counteract background disturbances and dramatic variations. Specifically, PGN is a video segmentation network to obtain segmentation mask proposals, while AEN is a network we specially designed to produce appearance-level embedding semantics for PGN, thereby unleashing the capability of PGN in VPS. Our AEN consists of a cross-scale region linking (CRL) module and a cross-wise scale alignment (CSA) module. The former screens reliable background information against background disturbances by constructing linking of region semantics, while the latter performs the scale alignment to resist dramatic variations by modeling the center-perceived motion dependence with a cross-wise manner. We further introduce a parameter-free semantic interaction to embed the semantics of AEN into PGN to obtain the segmentation results. Extensive experiments on CVC-612 and SUN-SEG demonstrate that our approach achieves better performance than other state-of-the-art methods. Codes are available at https://github.com/zhixue-fang/EUVPS.

Recent advancements in deep learning have greatly improved the efficiency of auxiliary medical diagnostics. However, concerns over patient privacy and data annotation costs restrict the viability of centralized training models. In response, federated semi-supervised learning has garnered substantial attention from medical institutions. However, it faces challenges arising from knowledge discrepancies among local clients and class imbalance in non-independent and identically distributed data. Existing methods like class balance adaptation for addressing class imbalance often overlook low-confidence yet valuable rare samples in unlabeled data and may compromise client privacy. To address these issues, we propose a novel framework with class awareness balance and dual teacher distillation called FedCD. FedCD introduces a global-local framework to balance and purify global and local knowledge. Additionally, we introduce a novel class awareness balance module to effectively explore potential rare classes and encourage balanced learning in unlabeled clients. Importantly, our approach prioritizes privacy protection by only exchanging network parameters during communication. Experimental results on two medical datasets under various settings demonstrate the effectiveness of FedCD. The code is available at https://github.com/YunzZ-Liu/FedCD.

Diffusion Models have shown remarkable performance in image generation tasks, which are capable of generating diverse and realistic image content. When adopting diffusion models for image restoration, the crucial challenge lies in how to preserve high-level image fidelity in the random-ness diffusion process and generate accurate background structures and realistic texture details. In this paper, we propose a general framework and develop a Diffusion Texture Prior Model (DTPM) for image restoration tasks. DTPM explicitly models high-quality texture details through the diffusion process, rather than global contextual content. In phase one of the training stage, we pretrain DTPM on approximately 55K high-quality image samples, after which we freeze most of its parameters. In phase two, we insert conditional guidance adapters into DTPM and equip it with an initial predictor, thereby facilitating its rapid adaptation to downstream image restoration tasks. Our DTPM could mitigate the randomness of traditional diffusion models by utilizing encapsulated rich and diverse texture knowledge and background structural information provided by the initial predictor during the sampling process.

We present a novel semi-supervised framework for breast ultrasound (BUS) image segmentation, which is a very challenging task owing to (1) large scale and shape variations of breast lesions and (2) extremely ambiguous boundaries caused by massive speckle noise and artifacts in BUS images. While existing models achieved certain progress in this task, we believe the main bottleneck nowadays for further improvement is that we still cannot deal with hard cases well. Our framework aims to break through this bottleneck, which includes two innovative components: an adaptive patch augmentation scheme and a hard-patch contrastive learning module. We first identify hard patches by computing the average entropy of each patch and then shield hard patches to prevent them from being cropped out while performing random patch cutmix. Such a scheme is able to prevent hard regions from being inadequately trained under strong augmentation. We further develop a new hard-patch contrastive learning algorithm to direct model attention to hard regions by applying extra contrast to pixels in hard patches, further improving segmentation performance on hard cases. We demonstrate the superior-ity of our framework to state-of-the-art approaches on two famous BUS datasets, achieving better performance under different labeling conditions. The code is available at https://github.com/jjjsyyy/PH-Net.

We present a novel semantic segmentation approach for incremental nuclei segmentation from histopathological images, which is a very challenging task as we have to in-crementally optimize existing models to make them perform well in both old and new classes without using training samples of old classes. Yet, it is an indispensable component of computer-aided diagnosis systems. The proposed approach has two key techniques. First, we propose a new future-class awareness mechanism by separating some potential regions for future classes from background based on their similari-ties to both old and new classes in the representation space. With this mechanism, we can not only reserve more parameter space for future updates but also enhance the repre-sentation capability of learned features. We further propose an innovative compatibility-inspired distillation scheme to make our model take full advantage of the knowledge learned by the old model. We conducted extensive experiments on two famous histopathological datasets and the results demonstrate the proposed approach achieves much better performance than state-of-the-art approaches. The code is available at https://github.com/why199911/nSeg.

Real-world vision tasks frequently suffer from the appearance of unexpected adverse weather conditions, including rain, haze, snow, and raindrops. In the last decade, convolutional neural networks and vision transformers have yielded outstanding results in single-weather video removal. However, due to the absence of appropriate adaptation, most of them fail to generalize to other weather conditions. Although ViWS-Net is proposed to remove ad-verse weather conditions in videos with a single set of pre-trained weights, it is seriously blinded by seen weather at train-time and degenerates when coming to unseen weather during test-time. In this work, we introduce test-time adaptation into adverse weather removal in videos, and propose the first framework that integrates test-time adaptation into the iterative diffusion reverse process. Specifically, we devise a diffusion-based network with a novel temporal noise model to efficiently explore frame-correlated information in degraded video clips at training stage. During inference stage, we introduce a proxy task named Diffusion Tubelet Self-Calibration to learn the primer distribution of test video stream and optimize the model by approx-imating the temporal noise model for online adaptation. Experimental results, on benchmark datasets, demonstrate that our Test-Time Adaptation method with Diffusion-based network(Diff- TTA) outperforms state-of-the-art methods in terms of restoring videos degraded by seen weather conditions. Its generalizable capability is validated with unseen weather conditions in synthesized and real-world videos.

Predicting multivariate time series is crucial, demanding precise modeling of intricate patterns, including inter-series dependencies and intra-series variations. Distinctive trend characteristics in each time series pose challenges, and existing methods, relying on basic moving average kernels, may struggle with the non-linear structure and complex trends in real-world data. Given that, we introduce a learnable decomposition strategy to capture dynamic trend information more reasonably. Additionally, we propose a dual attention module tailored to capture inter-series dependencies and intra-series variations simultaneously for better time series forecasting, which is implemented by channel-wise self-attention and autoregressive self-attention. To evaluate the effectiveness of our method, we conducted experiments across eight open-source datasets and compared it with the state-of-the-art methods. Through the comparison results, our $\textbf{Leddam}$ ($\textbf{LE}arnable$ $\textbf{D}ecomposition$ and $\textbf{D}ual $ $\textbf{A}ttention$ $\textbf{M}odule$) not only demonstrates significant advancements in predictive performance but also the proposed decomposition strategy can be plugged into other methods with a large performance-boosting, from 11.87% to 48.56% MSE error degradation. Code is available at this link: https://github.com/Levi-Ackman/Leddam.

Lumbar disc herniation, as one of the most common spinal degeneration diseases, significantly affects the quality of people’s lives. Effective identification and diagnosis of this disease is highly demanded and crucial to improve lumbar disc health care. In this paper, we propose a unified framework for diagnosing multiple lumbar degeneration diseases in MRI. Considering the basis of diagnosis is the accurate lumbar identification of vertebrae and discs, we thus tailor an anatomical knowledge-based process to identify the index of the detected vertebrae and discs. Specifically, the main difficulty of diagnosis lies in the accurate classification of the disc degenerative level if only one view of MRI is available. To combat this problem, we introduce multi-view and multi-scale MRI clues to the model learning, and equip our framework with a context-guided multi-view feature fusion module to fully exploit spatial-correlations and semantic-correlations in multi-view MRI, leading to significant improvements of diagnosis. Extensive results on two public datasets demonstrate the superiority of our proposed framework over the existing competitives in terms of lumbar localization, identification, and diagnosis.

Current point cloud denoising (PCD) models optimize single networks, trying to make their parameters adaptive to each point in a large pool of point clouds. Such a denoising network paradigm neglects that different points are often corrupted by different levels of noise and they may convey different geometric structures. Thus, the intricacy of both noise and geometry poses side effects including remnant noise, wrongly-smoothed edges, and distorted shape after denoising. We propose PathNet, a path-selective PCD paradigm based on reinforcement learning (RL). Unlike existing efforts, PathNet enables dynamic selection of the most appropriate denoising path for each point, best moving it onto its underlying surface. We have two more contributions besides the proposed framework of path-selective PCD for the first time. First, to leverage geometry expertise and benefit from training data, we propose a noise- and geometry-aware reward function to train the routing agent in RL. Second, the routing agent and the denoising network are trained jointly to avoid under- and over-smoothing. Extensive experiments show promising improvements of PathNet over its competitors, in terms of the effectiveness for removing different levels of noise and preserving multi-scale surface geometries. Furthermore, PathNet generalizes itself more smoothly to real scans than cutting-edge models.

Most state-of-the-art models for biomedical image segmentation are developed based on U-shape architecture, which has two renowned, yet mutually affected, shortcomings: 1) difficulties in capturing global long-range dependencies, and 2) semantic information dilution in the decoding process. In this paper, we propose a novel network with a new object-aware module (OAM) to effectively establish global dependencies at multiple levels within the network and compensate high-level semantic information dilution when fusing the extracted multi-level features; we call the network MOG-Net. Specifically, the OAM is designed to figure out the relations between each pixel and targeting object region and recalibrate class-level semantic information according to the relations. Compared with non-local models, which construct pixel-wise global dependencies, our OAM is more efficient and target-specific, enabling us to achieve satisfactory results with less extra computational overhead. In addition, we embed a pyramid context encoder module (PCEM) in the proposed OAM to alleviate semantic information dilution; this scheme is able to bridge the spatial-semantic gap when fusing features extracted from different levels. We extensively evaluate the proposed MOG-Net on four diverse biomedical image segmentation tasks with different imaging modalities, achieving segmentation performance with 88.19%, 90.95% and 66.03% in Dice on three one-class datasets, as well as 88.83% and 87.11% in Dice for two classes on a multi-class dataset, respectively. Experimental results demonstrate the effectiveness of the proposed method, consistently outperforming state-of-the-art methods in most evaluation metrics.Note to Practitioners—Semantic segmentation of biomedical images is a critical prerequisite for subsequent diagnosis, treatment, and quantitative tasks in clinical practice. This article proposes a novel biomedical image segmentation network, namely MOG-Net, with a new object-aware module (OAM) to model global context dependencies from a category perspective and a pyramid context encoder module (PCEM) to enhance feature representation capabilities of spatial and channel dimensions. We experimentally demonstrate the effectiveness and generalization capability of proposed MOG-Net on diverse biomedical image segmentation tasks with different imaging modalities. We believe that our proposed method can serve as a practical clinical tool and has the potential to be applied to existing computer-aided medical systems and clinical measurement.

Automatic kidney and tumor segmentation from CT volumes is a critical prerequisite/tool for diagnosis and surgical treatment (such as partial nephrectomy). However, it remains a particularly challenging issue as kidneys and tumors often exhibit large-scale variations, irregular shapes, and blurring boundaries. We propose a novel 3-D network to comprehensively tackle these problems; we call it 3DSN-Net. Compared with existing solutions, it has two compelling characteristics. First, with a new scale-aware feature extraction (SAFE) module, the proposed 3DSN-Net is capable of adaptively selecting appropriate receptive fields according to the sizes of targets instead of indiscriminately enlarging them, which is particularly essential for improving the segmentation accuracy of the tumor with large scale variation. Second, we propose a novel yet efficient nonlocal context guidance (NCG) mechanism to capture global dependencies to tackle irregular shapes and blurring boundaries of kidneys and tumors. Instead of directly harnessing a 3-D NCG mechanism, which makes the number of parameters exponentially increase and hence the network difficult to be trained under limited training data, we develop a 2.5D NCG mechanism based on projections of feature cubes, which achieves a tradeoff between segmentation accuracy and network complexity. We extensively evaluate the proposed 3DSN-Net on the famous KiTS dataset with many challenging kidney and tumor cases. Experimental results demonstrate our solution consistently outperforms state-of-the-art 3-D networks after being equipped with scale aware and NCG mechanisms, particularly for tumor segmentation.

Reversible face anonymization, unlike traditional face pixelization, seeks to replace sensitive identity information in facial images with synthesized alternatives, preserving privacy without sacrificing image clarity. Traditional methods, such as encoder-decoder networks, often result in significant loss of facial details due to their limited learning capacity. Additionally, relying on latent manipulation in pre-trained GANs can lead to changes in ID-irrelevant attributes, adversely affecting data utility due to GAN inversion inaccuracies. This paper introduces G2Face, which leverages both generative and geometric priors to enhance identity manipulation, achieving high-quality reversible face anonymization without compromising data utility. We utilize a 3D face model to extract geometric information from the input face, integrating it with a pre-trained GAN-based decoder. This synergy of generative and geometric priors allows the decoder to produce realistic anonymized faces with consistent geometry. Moreover, multi-scale facial features are extracted from the original face and combined with the decoder using our novel identity-aware feature fusion blocks (IFF). This integration enables precise blending of the generated facial patterns with the original ID-irrelevant features, resulting in accurate identity manipulation. Extensive experiments demonstrate that our method outperforms existing state-of-the-art techniques in face anonymization and recovery, while preserving high data utility. Code is available athttps://github.com/Harxis/G2Face.

While the wisdom of training an image dehazing model on synthetic hazy data can alleviate the difficulty of collecting real-world hazy/clean image pairs, it brings the well-known domain shift problem. From a different yet new perspective, this paper explores contrastive learning with an adversarial training effort to leverage unpaired real-world hazy and clean images, thus alleviating the domain shift problem and enhancing the network's generalization ability in real-world scenarios. We propose an effective unsupervised contrastive learning paradigm for image dehazing, dubbed UCL-Dehaze. Unpaired real-world clean and hazy images are easily captured, and will serve as the important positive and negative samples respectively when training our UCL-Dehaze network. To train the network more effectively, we formulate a new self-contrastive perceptual loss function, which encourages the restored images to approach the positive samples and keep away from the negative samples in the embedding space. Besides the overall network architecture of UCL-Dehaze, adversarial training is utilized to align the distributions between the positive samples and the dehazed images. Compared with recent image dehazing works, UCL-Dehaze does not require paired data during training and utilizes unpaired positive/negative data to better enhance the dehazing performance. We conduct comprehensive experiments to evaluate our UCL-Dehaze and demonstrate its superiority over the state-of-the-arts, even only 1,800 unpaired real-world images are used to train our network. Source code is publicly available at https://github.com/yz-wang/UCL-Dehaze.

Grading laryngeal squamous cell carcinoma (LSCC) based on histopathological images is a clinically significant yet challenging task. However, more low-effect background semantic information appeared in the feature maps, feature channels, and class activation maps, which caused a serious impact on the accuracy and interpretability of LSCC grading. While the traditional transformer block makes extensive use of parameter attention, the model overlearns the low-effect background semantic information, resulting in ineffectively reducing the proportion of background semantics. Therefore, we propose an end-to-end network with transformers constrained by learned-parameter-free attention (LA-ViT), which improve the ability to learn high-effect target semantic information and reduce the proportion of background semantics. Firstly, according to generalized linear model and probabilistic, we demonstrate that learned-parameter-free attention (LA) has a stronger ability to learn highly effective target semantic information than parameter attention. Secondly, the first-type LA transformer block of LA-ViT utilizes the feature map position subspace to realize the query. Then, it uses the feature channel subspace to realize the key, and adopts the average convergence to obtain a value. And those construct the LA mechanism. Thus, it reduces the proportion of background semantics in the feature maps and feature channels. Thirdly, the second-type LA transformer block of LA-ViT uses the model probability matrix information and decision level weight information to realize key and query, respectively. And those realize the LA mechanism. So, it reduces the proportion of background semantics in class activation maps. Finally, we build a new complex semantic LSCC pathology image dataset to address the problem, which is less research on LSCC grading models because of lacking clinically meaningful datasets. After extensive experiments, the whole metrics of LA-ViT outperform those of other state-of-the-art methods, and the visualization maps match better with the regions of interest in the pathologists' decision-making. Moreover, the experimental results conducted on a public LSCC pathology image dataset show that LA-ViT has superior generalization performance to that of other state-of-the-art methods.

Machine learning (ML), the stem of today's artificial intelligence, has shown significant growth in the field of biomedical and health informatics. On the one hand, ML techniques are becoming more complex in order to deal with real-world data. On the other hand, ML is also more and more accessible to broader users. For example, automated machine learning products are enabling users to build their own ML models without writing code [1].

Gene expression data can serve for analyzing the genes with changed expressions, the correlation between genes and the influence of different circumstance on gene activities. However, labeling a large number of gene expression data is laborious and time-consuming. The insufficient labeled data pose a challenge to construct the deep learning model. Currently, some graph neural networks (GNN) based on semi-supervised learning mechanism only focus on the feature space and sample space of gene expression data, possibly affecting the accuracy. This article puts forward a novel semi-supervised graph neural network model (SFWN). Firstly, we use the external knowledge of gene expression data for constructing a feature graph, a similarity kernel, and a sample graph for the first time. Later, a novel semi-supervised learning algorithm (SGA) is proposed to extract the data relationship and obtain the global sample structure better. A graph sparse module (SGCN) is also proposed to process sparse representation with gene expression data classification. To overcome the over smoothing problem, a new feature calculation method based on two spaces is proposed to feature representation analysis and calculation in this model. According to a lot of experiments and ablation studies conducted on several public datasets, SFWN exhibits a better effect and is superior to the state-of-the-art approaches (the accuracy and F1-Score are 0.9993 and 0.9899, respectively). Experimental results showed that the proposed SFWN model has strong gene expression feature learning and representation ability, and may provide a new insight and tool for relevant disease diagnosis and clinic practice.

Fusing multi-modal radiology and pathology data with complementary information can improve the accuracy of tumor typing. However, collecting pathology data is difficult since it is high-cost and sometimes only obtainable after the surgery, which limits the application of multi-modal methods in diagnosis. To address this problem, we propose comprehensively learning multi-modal radiology-pathology data in training, and only using uni-modal radiology data in testing. Concretely, a Memory-aware Hetero-modal Distillation Network (MHD-Net) is proposed, which can distill well-learned multi-modal knowledge with the assistance of memory from the teacher to the student. In the teacher, to tackle the challenge in hetero-modal feature fusion, we propose a novel spatial-differentiated hetero-modal fusion module (SHFM) that models spatial-specific tumor information correlations across modalities. As only radiology data is accessible to the student, we store pathology features in the proposed contrast-boosted typing memory module (CTMM) that achieves type-wise memory updating and stage-wise contrastive memory boosting to ensure the effectiveness and generalization of memory items. In the student, to improve the cross-modal distillation, we propose a multi-stage memory-aware distillation (MMD) scheme that reads memory-aware pathology features from CTMM to remedy missing modal-specific information. Furthermore, we construct a Radiology-Pathology Thymic Epithelial Tumor (RPTET) dataset containing paired CT and WSI images with annotations. Experiments on the RPTET and CPTAC-LUAD datasets demonstrate that MHD-Net significantly improves tumor typing and outperforms existing multi-modal methods on missing modality situations.

Multimodal neuroimaging provides complementary information critical for accurate early diagnosis of Alzheimer's disease (AD). However, the inherent variability between multimodal neuroimages hinders the effective fusion of multimodal features. Moreover, achieving reliable and interpretable diagnoses in the field of multimodal fusion remains challenging. To address them, we propose a novel multimodal diagnosis network based on multi-fusion and disease-induced learning (MDL-Net) to enhance early AD diagnosis by efficiently fusing multimodal data. Specifically, MDL-Net proposes a multi-fusion joint learning (MJL) module, which effectively fuses multimodal features and enhances the feature representation from global, local, and latent learning perspectives. MJL consists of three modules, global-aware learning (GAL), local-aware learning (LAL), and outer latent-space learning (LSL) modules. GAL via a self-adaptive Transformer (SAT) learns the global relationships among the modalities. LAL constructs local-aware convolution to learn the local associations. LSL module introduces latent information through outer product operation to further enhance feature representation. MDL-Net integrates the disease-induced region-aware learning (DRL) module via gradient weight to enhance interpretability, which iteratively learns weight matrices to identify AD-related brain regions. We conduct the extensive experiments on public datasets and the results confirm the superiority of our proposed method. Our code will be available at: https://github.com/qzf0320/MDL-Net.

Surgical instrument segmentation is fundamentally important for facilitating cognitive intelligence in robot-assisted surgery. Although existing methods have achieved accurate instrument segmentation results, they simultaneously generate segmentation masks of all instruments, which lack the capability to specify a target object and allow an interactive experience. This paper focuses on a novel and essential task in robotic surgery, i.e., Referring Surgical Video Instrument Segmentation (RSVIS), which aims to automatically identify and segment the target surgical instruments from each video frame, referred by a given language expression. This interactive feature offers enhanced user engagement and customized experiences, greatly benefiting the development of the next generation of surgical education systems. To achieve this, this paper constructs two surgery video datasets to promote the RSVIS research. Then, we devise a novel Video-Instrument Synergistic Network (VIS-Net) to learn both video-level and instrument-level knowledge to boost performance, while previous work only utilized video-level information. Meanwhile, we design a Graph-based Relation-aware Module (GRM) to model the correlation between multi-modal information (i.e., textual description and video frame) to facilitate the extraction of instrument-level information. Extensive experimental results on two RSVIS datasets exhibit that the VIS-Net can significantly outperform existing state-of-the-art referring segmentation methods. We will release our code and dataset for future research (https://github.com/whq-xxh/RSVIS).

Automated classification of breast cancer subtypes from digital pathology images has been an extremely challenging task due to the complicated spatial patterns of cells in the tissue micro-environment. While newly proposed graph transformers are able to capture more long-range dependencies to enhance accuracy, they largely ignore the topological connectivity between graph nodes, which is nevertheless critical to extract more representative features to address this difficult task. In this paper, we propose a novel connectivity-aware graph transformer (CGT) for phenotyping the topology connectivity of the tissue graph constructed from digital pathology images for breast cancer classification. Our CGT seamlessly integrates connectivity embedding to node feature at every graph transformer layer by using local connectivity aggregation, in order to yield more comprehensive graph representations to distinguish different breast cancer subtypes. In light of the realistic intercellular communication mode, we then encode the spatial distance between two arbitrary nodes as connectivity bias in self-attention calculation, thereby allowing the CGT to distinctively harness the connectivity embedding based on the distance of two nodes. We extensively evaluate the proposed CGT on a large cohort of breast carcinoma digital pathology images stained by Haematoxylin & Eosin. Experimental results demonstrate the effectiveness of our CGT, which outperforms state-of-the-art methods by a large margin. Codes are released on https://github.com/wang-kang-6/CGT.

Taking advantage of multi-modal radiology-pathology data with complementary clinical information for cancer grading is helpful for doctors to improve diagnosis efficiency and accuracy. However, radiology and pathology data have distinct acquisition difficulties and costs, which leads to incomplete-modality data being common in applications. In this work, we propose a Memory- and Gradient-guided Incomplete Modal-modal Learning (MGIML) framework for cancer grading with incomplete radiology-pathology data. Firstly, to remedy missing-modality information, we propose a Memory-driven Hetero-modality Complement (MH-Complete) scheme, which constructs modal-specific memory banks constrained by a coarse-grained memory boosting (CMB) loss to record generic radiology and pathology feature patterns, and develops a cross-modal memory reading strategy enhanced by a fine-grained memory consistency (FMC) loss to take missing-modality information from well-stored memories. Secondly, as gradient conflicts exist between missing-modality situations, we propose a Rotation-driven Gradient Homogenization (RG-Homogenize) scheme, which estimates instance-specific rotation matrices to smoothly change the feature-level gradient directions, and computes confidence-guided homogenization weights to dynamically balance gradient magnitudes. By simultaneously mitigating gradient direction and magnitude conflicts, this scheme well avoids the negative transfer and optimization imbalance problems. Extensive experiments on CPTAC-UCEC and CPTAC-PDA datasets show that the proposed MGIML framework performs favorably against state-of-the-art multi-modal methods on missing-modality situations.

Existing federated learning works mainly focus on the fully supervised training setting. In realistic scenarios, however, most clinical sites can only provide data without annotations due to the lack of resources or expertise. In this work, we are concerned with the practical yet challenging federated semi-supervised segmentation (FSSS), where labeled data are only with several clients and other clients can just provide unlabeled data. We take an early attempt to tackle this problem and propose a novel FSSS method with prototype-based pseudo-labeling and contrastive learning. First, we transmit a labeled-aggregated model, which is obtained based on prototype similarity, to each unlabeled client, to work together with the global model for debiased pseudo labels generation via a consistency- and entropy-aware selection strategy. Second, we transfer image-level prototypes from labeled datasets to unlabeled clients and conduct prototypical contrastive learning on unlabeled models to enhance their discriminative power. Finally, we perform the dynamic model aggregation with a designed consistency-aware aggregation strategy to dynamically adjust the aggregation weights of each local model. We evaluate our method on COVID-19 X-ray infected region segmentation, COVID-19 CT infected region segmentation and colorectal polyp segmentation, and experimental results consistently demonstrate the effectiveness of our proposed method. Codes areavailable at https://github.com/zhangbaiming/FedSemiSeg.

Incomplete multi-view clustering (IMVC) aims to leverage complementary information from multi-view data with missing instances to enhance clustering performance. Many existing IMVC methods exhibit limitations in effectively exploiting hidden information and addressing distribution differences between views and modules. To address these challenges, we present a novel IMVC framework that leverages the proposed stack feature-based matrix completion to impute the missing instances, enhancing the exploitation of underlying information. We also incorporate graph consensus to integrate graph structures learned from both completed and observed data. Additionally, we introduce correntropy-induced metric as a flexible measurement to adaptively assign different constraints to various views and modules. Furthermore, we derive an efficient iterative algorithm based on Fenchel conjugate and accelerated block coordinate update (BCU) to solve the joint learning problem. Experimental results on eight benchmark datasets demonstrate the superior performance of our method compared to state-of-the-art IMVC methods across various metrics.

The emerging matrix learning methods have achieved promising performances in electroencephalogram (EEG) classification by exploiting the structural information between the columns or rows of feature matrices. Due to the intersubject variability of EEG data, these methods generally need to collect a large amount of labeled individual EEG data, which would cause fatigue and inconvenience to the subjects. Insufficient subject-specific EEG data will weaken the generalization capability of the matrix learning methods in neural pattern decoding. To overcome this dilemma, we propose an adaptive multimodel knowledge transfer matrix machine (AMK-TMM), which can selectively leverage model knowledge from multiple source subjects and capture the structural information of the corresponding EEG feature matrices. Specifically, by incorporating least-squares (LS) loss with spectral elastic net regularization, we first present an LS support matrix machine (LS-SMM) to model the EEG feature matrices. To boost the generalization capability of LS-SMM in scenarios with limited EEG data, we then propose a multimodel adaption method, which can adaptively choose multiple correlated source model knowledge with a leave-one-out cross-validation strategy on the available target training data. We extensively evaluate our method on three independent EEG datasets. Experimental results demonstrate that our method achieves promising performances on EEG classification.

Magnetic resonance imaging (MRI) possesses the unique versatility to acquire images under a diverse array of distinct tissue contrasts, which makes multicontrast super-resolution (SR) techniques possible and needful. Compared with single-contrast MRI SR, multicontrast SR is expected to produce higher quality images by exploiting a variety of complementary information embedded in different imaging contrasts. However, existing approaches still have two shortcomings: 1) most of them are convolution-based methods and, hence, weak in capturing long-range dependencies, which are essential for MR images with complicated anatomical patterns and 2) they ignore to make full use of the multicontrast features at different scales and lack effective modules to match and aggregate these features for faithful SR. To address these issues, we develop a novel multicontrast MRI SR network via transformer-empowered multiscale feature matching and aggregation, dubbed McMRSR$^{++}$. First, we tame transformers to model long-range dependencies in both reference and target images at different scales. Then, a novel multiscale feature matching and aggregation method is proposed to transfer corresponding contexts from reference features at different scales to the target features and interactively aggregate them Furthermore, a texture-preserving branch and a contrastive constraint are incorporated into our framework for enhancing the textural details in the SR images. Experimental results on both public and clinical in vivo datasets show that McMRSR$^{++}$outperforms state-of-the-art methods under peak signal to noise ratio (PSNR), structure similarity index measure (SSIM), and root mean square error (RMSE) metrics significantly. Visual results demonstrate the superiority of our method in restoring structures, demonstrating its great potential to improve scan efficiency in clinical practice.

Deep neural networks often suffer from performance inconsistency for multiorgan segmentation in medical images; some organs are segmented far worse than others. The main reason might be organs with different levels of learning difficulty for segmentation mapping, due to variations such as size, texture complexity, shape irregularity, and imaging quality. In this article, we propose a principled class-reweighting algorithm, termed dynamic loss weighting, which dynamically assigns a larger loss weight to organs if they are discriminated as more difficult to learn according to the data and network's status, for forcing the network to learn from them more to maximally promote the performance consistency. This new algorithm uses an extra autoencoder to measure the discrepancy between the segmentation network's output and the ground truth and dynamically estimates the loss weight of organs per the contribution of the organ to the new updated discrepancy. It can capture the variation in organs' learning difficult during training, and it is neither sensitive to data's property nor dependent on human priors. We evaluate this algorithm in two multiorgan segmentation tasks: abdominal organs and head-neck structures, on publicly available datasets, with positive results obtained from extensive experiments which confirm the validity and effectiveness. Source codes are available at: https://github.com/YouyiSong/Dynamic-Loss-Weighting.

Sentiment and emotion, which correspond to long-term and short-lived human feelings, are closely linked to each other, leading to the fact that sentiment analysis and emotion recognition are also two interdependent tasks in natural language processing (NLP). One task often leverages the shared knowledge from another task and performs better when solved in a joint learning paradigm. Conversational context dependency, multi-modal interaction, and multi-task correlation are three key factors that contribute to this joint paradigm. However, none of the recent approaches have considered them in a unified framework. To fill this gap, we propose a multi-modal, multi-task interactive graph attention network, termed M3GAT, to simultaneously solve the three problems. At the heart of the model is a proposed interactive conversation graph layer containing three core sub-modules, which are: (1) local-global context connection for modeling both local and global conversational context, (2) cross-modal connection for learning multi-modal complementary and (3) cross-task connection for capturing the correlation across two tasks. Comprehensive experiments on three benchmarking datasets, MELD, MEISD, and MSED, show the effectiveness of M3GAT over state-of-the-art baselines with the margin of 1.88%, 5.37%, and 0.19% for sentiment analysis, and 1.99%, 3.65%, and 0.13% for emotion recognition, respectively. In addition, we also show the superiority of multi-task learning over the single-task framework.

Exploiting long-range semantic contexts and geometric information is crucial to infer salient objects from RGB and depth features. However, existing methods mainly focus on excavating local features within fixed regions by continuously feeding forward networks. In this article, we introduce Dynamic Message Propagation (DMP) to dynamically learn context information within more flexible regions. We integrate DMP into a Siamese-based network to process the RGB image and depth map separately and design a multi-level feature fusion module to explore cross-level information between refined RGB and depth features. Extensive experiments show clear improvements of our method over 17 methods on six benchmark datasets for RGB-D salient object detection (SOD). Additionally, our method outperforms its competitors for the video SOD task. Code is available at https://github.com/chenbaian-cs/DMPNet .

There is a prevailing trend towards fusing multi-modal information for 3D object detection (3OD). However, challenges related to computational efficiency, plug-and-play capabilities, and accurate feature alignment have not been adequately addressed in the design of multi-modal fusion networks. In this paper, we present PointSee, a lightweight, flexible, and effective multi-modal fusion solution to facilitate various 3OD networks by semantic feature enhancement of point clouds (e.g., LiDAR or RGB-D data) assembled with scene images. Beyond the existing wisdom of 3OD, PointSee consists of a hidden module (HM) and a seen module (SM): HM decorates point clouds using 2D image information in an offline fusion manner, leading to minimal or even no adaptations of existing 3OD networks; SM further enriches the point clouds by acquiring point-wise representative semantic features, leading to enhanced performance of existing 3OD networks. Besides the new architecture of PointSee, we propose a simple yet efficient training strategy, to ease the potential inaccurate regressions of 2D object detection networks. Extensive experiments on the popular outdoor/indoor benchmarks show quantitative and qualitative improvements of our PointSee over thirty-five state-of-the-art methods.

How will you repair a physical object with some missings? You may imagine its original shape from previously captured images, recover its overall (global) but coarse shape first, and then refine its local details. We are motivated to imitate the physical repair procedure to address point cloud completion. To this end, we propose a cross-modal shape-transfer dual-refinement network (termed CSDN), a coarse-to-fine paradigm with images of full-cycle participation, for quality point cloud completion. CSDN mainly consists of "shape fusion" and "dual-refinement" modules to tackle the cross-modal challenge. The first module transfers the intrinsic shape characteristics from single images to guide the geometry generation of the missing regions of point clouds, in which we propose IPAdaIN to embed the global features of both the image and the partial point cloud into completion. The second module refines the coarse output by adjusting the positions of the generated points, where the local refinement unit exploits the geometric relation between the novel and the input points by graph convolution, and the global constraint unit utilizes the input image to fine-tune the generated offset. Different from most existing approaches, CSDN not only explores the complementary information from images but also effectively exploits cross-modal data in the whole coarse-to-fine completion procedure. Experimental results indicate that CSDN performs favorably against twelve competitors on the cross-modal benchmark.

Automatic segmentation of left ventricular endocardium in echocardiography videos is critical for assessing various cardiac functions and improving the diagnosis of cardiac diseases. It is yet a challenging task due to heavy speckle noise, significant shape variability of cardiac structure, and limited labeled data. Particularly, the real-time demand in clinical practice makes this task even harder. In this paper, we propose a novel proxy- and kernel-based semi-supervised segmentation network (PKEcho-Net) to comprehensively address these challenges. We first propose a multi-scale region proxy (MRP) mechanism to model the region-wise contexts, in which a learnable region proxy with an arbitrary shape is developed in each layer of the encoder, allowing the network to identify homogeneous semantics and hence alleviate the influence of speckle noise on segmentation. To sufficiently and efficiently exploit temporal consistency, different from traditional methods which only utilize the temporal contexts of two neighboring frames via feature warping or self-attention mechanism, we formulate the semi-supervised segmentation with a group of learnable kernels, which can naturally and uniformly encode the appearances of left ventricular endocardium, as well as extracting the inter-frame contexts across the whole video to resist the fast shape variability of cardiac structures. Extensive experiments have been conducted on two famous public echocardiography video datasets, EchoNet-Dynamic and CAMUS. Our model achieves the best performance-efficiency trade-off when compared with other state-of-the-art approaches, attaining comparative accuracy with a much faster speed. The code is available at https://github.com/JingyinLin/PKEcho-Net.

With sufficient centralized training data coming from multiple subjects, deep learning methods have achieved powerful EEG decoding performance. However, sending each individuals’ EEG data directly to a centralized server might cause privacy leakage. To overcome this issue, we present an inter-subject structure matching-based federated EEG decoding (FedEEG) framework. First, we introduce a center loss to each client (subject), which can learn multiple virtual class centers by averaging the corresponding class-specific EEG features. To mitigate the client drift issue, we then explicitly connect the learning across multiple clients by aligning their corresponding virtual class centers, thus helping to correct the local training for individual subject. The proposed FedEEG can promote the discriminative feature learning while preventing the privacy leakage issue. The experimental results on benchmark EEG datasets show that FedEEG outperforms state-of-the-art federated learning methods.

One critical challenge in 6D object pose estimation from a single RGBD image is efficient integration of two different modalities, i.e., color and depth. In this work, we tackle this problem by a novel Deep Fusion Transformer (DFTr) block that can aggregate cross-modality features for improving pose estimation. Unlike existing fusion methods, the proposed DFTr can better model cross-modality semantic correlation by leveraging their semantic similarity, such that globally enhanced features from different modalities can be better integrated for improved information extraction. Moreover, to further improve robustness and efficiency, we introduce a novel weighted vector-wise voting algorithm that employs a non-iterative global optimization strategy for precise 3D keypoint localization while achieving near real-time inference. Extensive experiments show the effectiveness and strong generalization capability of our proposed 3D keypoint voting algorithm. Results on four widely used benchmarks also demonstrate that our method outperforms the state-of-the-art methods by large margins. Code is available at https://github.com/junzastar/DFTr_Voting.

In this paper, we propose a novel network, SVDFormer, to tackle two specific challenges in point cloud completion: understanding faithful global shapes from incomplete point clouds and generating high-accuracy local structures. Current methods either perceive shape patterns using only 3D coordinates or import extra images with well-calibrated intrinsic parameters to guide the geometry estimation of the missing parts. However, these approaches do not always fully leverage the cross-modal self-structures available for accurate and high-quality point cloud completion. To this end, we first design a Self-view Fusion Network that leverages multiple-view depth image information to observe incomplete self-shape and generate a compact global shape. To reveal highly detailed structures, we then introduce a refinement module, called Self-structure Dual-generator, in which we incorporate learned shape priors and geometric self-similarities for producing new points. By perceiving the incompleteness of each point, the dual-path design disentangles refinement strategies conditioned on the structural type of each point. SVDFormer absorbs the wisdom of self-structures, avoiding any additional paired information such as color images with precisely calibrated camera intrinsic parameters. Comprehensive experiments indicate that our method achieves state-of-the-art performance on widely-used benchmarks. Code is available at https://github.com/czvvd/SVDFormer.

Non-photorealistic videos are in demand with the wave of the metaverse, but lack of sufficient research studies. This work aims to take a step forward to understand how humans perceive non-photorealistic videos with eye fixation (i.e., saliency detection), which is critical for enhancing media production, artistic design, and game user experience. To fill in the gap of missing a suitable dataset for this research line, we present NPF-200, the first large-scale multi-modal dataset of purely non-photorealistic videos with eye fixations. Our dataset has three characteristics: 1) it contains soundtracks that are essential according to vision and psychological studies; 2) it includes diverse semantic content and videos are of high-quality; 3) it has rich motions across and within videos. We conduct a series of analyses to gain deeper insights into this task and compare several state-of-the-art methods to explore the gap between natural images and non-photorealistic data. Additionally, as the human attention system tends to extract visual and audio features with different frequencies, we propose a universal frequency-aware multi-modal non-photorealistic saliency detection model called NPSNet, demonstrating the state-of-the-art performance of our task. The results uncover strengths and weaknesses of multi-modal network design and multi-domain training, opening up promising directions for future works. Our dataset and code can be found at https://github.com/Yangziyu/NPF200

Human communication has a multi-modal and multi-affection nature. The inter-relatedness of different emotions and sentiments poses a challenge to jointly detect multiple human affections with multi-modal clues. Recent advances in this field employed multi-task learning paradigms to render the inter-relatedness across tasks, but the scarcity of publicly available resources sets a limit to the potential of works. To fill this gap, we build the first Chinese Multi-modal Multi-Affection conversation (CMMA) dataset, which contains 3,000 multi-party conversations and 21,795 multi-modal utterances collected from various styles of TV-series. CMMA contains a wide variety of affection labels, including sentiment, emotion, sarcasm and humor, as well as the novel inter-correlations values between certain pairs of tasks. Moreover, it provides the topic and speaker information in conversations, which promotes better modeling of conversational context. On the dataset, we empirically analyze the influence of different data modalities and conversational contexts on different affection analysis tasks, and exhibit the practical benefit of inter-task correlations. The full dataset will be publicly available for research\footnote{https://github.com/annoymity2022/Chinese-Dataset}

Abstract Corona Virus disease 2019 (COVID‐19) has shattered people's daily lives and is spreading rapidly across the globe. Existing non‐pharmaceutical intervention solutions often require timely and precise selection of small areas of people for containment or even isolation. Although such containment has been successful in stopping or mitigating the spread of COVID‐19 in some countries, it has been criticized as inefficient or ineffective, because of the time‐delayed and sophisticated nature of the statistics on determining cases. To address these concerns, we propose a GSA‐ELM model based on a gravitational search algorithm to forecast the global number of active cases of COVID‐19. The model employs the gravitational search algorithm, which utilises the gravitational law between two particles to guide the motion of each particle to optimise the search for the global optimal solution, and utilises an extreme learning machine to address the effects of nonlinearity in the number of active cases. Extensive experiments are conducted on the statistical COVID‐19 dataset from Johns Hopkins University, the MAPE of the authors’ model is 7.79%, which corroborates the superiority of the model to state‐of‐the‐art methods.

Object detection remains challenging in hazy weather conditions due to the poor visibility of captured images. There are currently two types of detectors capable of adapting to varying weather conditions: (i) low‐level adaptation methods that combine one detector with an additional dehazing network and (ii) high‐level adaptation methods that explore various kinds of domain adaptation knowledge. However, neither of these approaches can achieve desirable performance due to their inherent limitations. We raise an intriguing question—if combining both low‐level adaptation and high‐level adaptation, can improve the generalization ability of a detector in hazy weather conditions? To answer it, we propose a Joint High‐Low Adaptation Object Detection paradigm (HLA‐HOD) in hazy weather conditions. By combining both low‐level adaptation and high‐level adaptation, HLA‐HOD achieves superior performance on hazy images without requiring ground‐truth bounding boxes or clean images. Extensive experiments demonstrate that our method outperforms state‐of‐the‐art low‐level and high‐level adaptation methods by a large margin both quantitatively and qualitatively.

Traditional monocular depth estimation assumes that all objects are reliably visible in the RGB color domain. However, this is not always the case as more and more buildings are decorated with transparent glass walls. This problem has not been explored due to the difficulties in annotating the depth levels of glass walls, as commercial depth sensors cannot provide correct feedbacks on transparent objects. Furthermore, estimating depths from transparent glass walls requires the aids of surrounding context, which has not been considered in prior works. To cope with this problem, we introduce the first Glass Walls Depth Dataset (GW-Depth dataset). We annotate the depth levels of transparent glass walls by propagating the context depth values within neighboring flat areas, and the glass segmentation mask and instance level line segments of glass edges are also provided. On the other hand, a tailored monocular depth estimation method is proposed to fully activate the glass wall contextual understanding. First, we propose to exploit the glass structure context by incorporating the structural prior knowledge embedded in glass boundary line segment detections. Furthermore, to make our method adaptive to scenes without structure context where the glass boundary is either absent in the image or too narrow to be recognized, we propose to derive a reflection context by utilizing the depth reliable points sampled according to the variance between two depth estimations from different resolutions. High-resolution depth is thus estimated by the weighted summation of depths by those reliable points. Extensive experiments are conducted to evaluate the effectiveness of the proposed dual context design. Superior performances of our method is also demonstrated by comparing with state-of-the-art methods. We present the first feasible solution for monocular depth estimation in the presence of glass walls, which can be widely adopted in autonomous navigation.

Labeling is onerous for crowd counting as it should annotate each individual in crowd images. Recently, several methods have been proposed for semi-supervised crowd counting to reduce the labeling efforts. Given a limited labeling budget, they typically select a few crowd images and densely label all individuals in each of them. Despite the promising results, we argue the None-or-All labeling strategy is suboptimal as the densely labeled individuals in each crowd image usually appear similar while the massive unlabeled crowd images may contain entirely diverse individuals. To this end, we propose to break the labeling chain of previous methods and make the first attempt to reduce spatial labeling redundancy for semi-supervised crowd counting. First, instead of annotating all the regions in each crowd image, we propose to annotate the representative ones only. We analyze the region representativeness from both vertical and horizontal directions of initially estimated density maps, and formulate them as cluster centers of Gaussian Mixture Models. Additionally, to leverage the rich unlabeled regions, we exploit the similarities among individuals in each crowd image to directly supervise the unlabeled regions via feature propagation instead of the error-prone label propagation employed in the previous methods. In this way, we can transfer the original spatial labeling redundancy caused by individual similarities to effective supervision signals on the unlabeled regions. Extensive experiments on the widely-used benchmarks demonstrate that our method can outperform previous best approaches by a large margin.

Convolution on 3D point clouds is widely researched yet far from perfect in geometric deep learning. The traditional wisdom of convolution characterises feature correspondences indistinguishably among 3D points, arising an intrinsic limitation of poor distinctive feature learning. In this article, we propose Adaptive Graph Convolution (AGConv) for wide applications of point cloud analysis. AGConv generates adaptive kernels for points according to their dynamically learned features. Compared with the solution of using fixed/isotropic kernels, AGConv improves the flexibility of point cloud convolutions, effectively and precisely capturing the diverse relations between points from different semantic parts. Unlike the popular attentional weight schemes, AGConv implements the adaptiveness inside the convolution operation instead of simply assigning different weights to the neighboring points. Extensive evaluations clearly show that our method outperforms state-of-the-arts of point cloud classification and segmentation on various benchmark datasets. Meanwhile, AGConv can flexibly serve more point cloud analysis approaches to boost their performance. To validate its flexibility and effectiveness, we explore AGConv-based paradigms of completion, denoising, upsampling, registration and circle extraction, which are comparable or even superior to their competitors.

Point normal, as an intrinsic geometric property of 3D objects, not only serves conventional geometric tasks such as surface consolidation and reconstruction, but also facilitates cutting-edge learning-based techniques for shape analysis and generation. In this paper, we propose a normal refinement network, called Refine-Net, to predict accurate normals for noisy point clouds. Traditional normal estimation wisdom heavily depends on priors such as surface shapes or noise distributions, while learning-based solutions settle for single types of hand-crafted features. Differently, our network is designed to refine the initial normal of each point by extracting additional information from multiple feature representations. To this end, several feature modules are developed and incorporated into Refine-Net by a novel connection module. Besides the overall network architecture of Refine-Net, we propose a new multi-scale fitting patch selection scheme for the initial normal estimation, by absorbing geometry domain knowledge. Also, Refine-Net is a generic normal estimation framework: 1) point normals obtained from other methods can be further refined, and 2) any feature module related to the surface geometric structures can be potentially integrated into the framework. Qualitative and quantitative evaluations demonstrate the clear superiority of Refine-Net over the state-of-the-arts on both synthetic and real-scanned datasets.

Printed photographs can be easily warped, wrinkled, and even deteriorated over time. Existing methods treat the restoration of scratches as a pure inpainting problem that neglects the underlying corrupted contextual knowledge. However, important underlying contents are hidden behind the scratches, which are essential hints for producing a semantically consistent result. Motivated by this insight, we explore how to harmonize the scratch-free features and noisy but essential scratch features to produce a visually consistent restoration. Specifically, in this paper, we propose an automatic retouching approach for scratched photographs with the aid of scratch/background context. We explicitly process scratch and background context in two stages. In the first stage, we mainly extract global scratch features, while the mask is introduced in the second stage to filter out and inpaint the scratches. Both contexts are carefully reciprocated for a faithful restoration. Particularly, we propose a Scratch Contextual Assisted Module (SCAM) to adaptively learn texture within the detected mask. This module utilizes the distance between the scratch mask-out feature and scratch encoder feature for modeling the pixel-wise correspondence, which determines the importance of the encoder feature within the scratch mask. Furthermore, to facilitate the evaluation of scratch restoration methods, we create two new scratched photo datasets which have 238 scratch/scratch-free photo pairs to promote the development in the scratch restoration field, namely Old Scratched Photo Dataset (OSPD) and Modern Scratched Photo Dataset (MSPD). Extensive experimental results on the proposed datasets demonstrate that our model outperforms existing methods. To extend the application, we also perform the proposed method on video samples and obtain visual-pleasing results. The code can be found athttps://github.com/cwyyt/Contextual-assisted-Scratched-Photo-Restoration.

Automatic polyp segmentation from colonoscopy videos is a prerequisite for the development of a computer-assisted colon cancer examination and diagnosis system. However, it remains a very challenging task owing to the large variation of polyps, the low contrast between polyps and background, and the blurring boundaries of polyps. More importantly, real-time performance is a necessity of this task, as it is anticipated that the segmented results can be immediately presented to the doctor during the colonoscopy intervention for his/her prompt decision and action. It is difficult to develop a model with powerful representation capability, yielding satisfactory segmentation results and, simultaneously, maintaining real-time performance. In this article, we present a novel lightweight context-aware network, namely, PolypSeg+, attempting to capture distinguishable features of polyps without increasing network complexity and sacrificing time performance. To achieve this, a set of novel lightweight techniques is developed and integrated into the proposed PolypSeg+, including an adaptive scale context (ASC) module equipped with a lightweight attention mechanism to tackle the large-scale variation of polyps, an efficient global context (EGC) module to promote the fusion of low-level and high-level features by excluding background noise and preserving boundary details, and a lightweight feature pyramid fusion (FPF) module to further refine the features extracted from the ASC and EGC. We extensively evaluate the proposed PolypSeg+ on two famous public available datasets for the polyp segmentation task: 1) Kvasir-SEG and 2) CVC-Endoscenestill. The experimental results demonstrate that our PolypSeg+ consistently outperforms other state-of-the-art networks by achieving better segmentation accuracy in much less running time. The code is available at https://github.com/szu-zzb/polypsegplus.

LiDAR and camera, as two different sensors, supply geometric (point clouds) and semantic (RGB images) information of 3-D scenes. However, it is still challenging for existing methods to fuse data from the two cross sensors, making them complementary for quality 3-D object detection (3OD). We propose ImLiDAR, a new 3OD paradigm to narrow the cross-sensor discrepancies by progressively fusing the multiscale features of camera Images and LiDAR point clouds. ImLiDAR enables to provide the detection head with cross-sensor yet robustly fused features. To achieve this, two core designs exist in ImLiDAR. First, we propose a cross-sensor dynamic message propagation (CDMP) module to combine the best of the multiscale image and point features. Second, we raise a direct set prediction problem that allows designing an effective set-based detector (SD) to tackle the inconsistency of the classification and localization confidences, and the sensitivity of hand-tuned hyperparameters. Besides, the novel SD can be detachable and easily integrated into various detection networks. Comparisons on the KITTI, nuScenes, and SUN-RGBD datasets all show clear visual and numerical improvements of our ImLiDAR over 45 state-of-the-art 3OD methods.

Cancer survival prediction requires exploiting related multimodal information (e.g., pathological, clinical and genomic features, etc.) and it is even more challenging in clinical practices due to the incompleteness of patient's multimodal data. Furthermore, existing methods lack sufficient intra- and inter-modal interactions, and suffer from significant performance degradation caused by missing modalities. This manuscript proposes a novel hybrid graph convolutional network, entitled HGCN, which is equipped with an online masked autoencoder paradigm for robust multimodal cancer survival prediction. Particularly, we pioneer modeling the patient's multimodal data into flexible and interpretable multimodal graphs with modality-specific preprocessing. HGCN integrates the advantages of graph convolutional networks (GCNs) and a hypergraph convolutional network (HCN) through node message passing and a hyperedge mixing mechanism to facilitate intra-modal and inter-modal interactions between multimodal graphs. With HGCN, the potential for multimodal data to create more reliable predictions of patient's survival risk is dramatically increased compared to prior methods. Most importantly, to compensate for missing patient modalities in clinical scenarios, we incorporated an online masked autoencoder paradigm into HGCN, which can effectively capture intrinsic dependence between modalities and seamlessly generate missing hyperedges for model inference. Extensive experiments and analysis on six cancer cohorts from TCGA show that our method significantly outperforms the state-of-the-arts in both complete and missing modal settings. Our codes are made available at https://github.com/lin-lcx/HGCN.

The tumor grading of laryngeal cancer pathological images needs to be accurate and interpretable. The deep learning model based on the attention mechanism-integrated convolution (AMC) block has good inductive bias capability but poor interpretability, whereas the deep learning model based on the vision transformer (ViT) block has good interpretability but weak inductive bias ability. Therefore, we propose an end-to-end ViT-AMC network (ViT-AMCNet) with adaptive model fusion and multiobjective optimization that integrates and fuses the ViT and AMC blocks. However, existing model fusion methods often have negative fusion: 1). There is no guarantee that the ViT and AMC blocks will simultaneously have good feature representation capability. 2). The difference in feature representations learning between the ViT and AMC blocks is not obvious, so there is much redundant information in the two feature representations. Accordingly, we first prove the feasibility of fusing the ViT and AMC blocks based on Hoeffding's inequality. Then, we propose a multiobjective optimization method to solve the problem that ViT and AMC blocks cannot simultaneously have good feature representation. Finally, an adaptive model fusion method integrating the metrics block and the fusion block is proposed to increase the differences between feature representations and improve the deredundancy capability. Our methods improve the fusion ability of ViT-AMCNet, and experimental results demonstrate that ViT-AMCNet significantly outperforms state-of-the-art methods. Importantly, the visualized interpretive maps are closer to the region of interest of concern by pathologists, and the generalization ability is also excellent. Our code is publicly available at https://github.com/Baron-Huang/ViT-AMCNet.

Collecting sufficient high-quality training data for deep neural networks is often expensive or even unaffordable in medical image segmentation tasks. We thus propose to train the network by using external data that can be collected in a cheaper way, e.g., crowd-sourcing. We show that by data discernment, the network is able to mine valuable knowledge from external data, even though the data distribution is very different from that of the original (internal) data. We discern the external data by learning an importance weight for each of them, with the goal to enhance the contribution of informative external data to network updating, while suppressing the data that are 'useless' or even 'harmful'. An iterative algorithm that alternatively estimates the importance weight and updates the network is developed by formulating the data discernment as a constrained nonlinear programming problem. It estimates the importance weight according to the distribution discrepancy between the external data and the internal dataset, and imposes a constraint to drive the network to learn more effectively, compared with the network without using the external data. We evaluate the proposed algorithm on two tasks: abdominal CT image and cervical smear image segmentation, using totally 6 publicly available datasets. The effectiveness of the algorithm is demonstrated by extensive experiments. Source codes are available at: https://github.com/YouyiSong/Data-Discernment.

Skin lesion segmentation from dermoscopy images is of great significance in the quantitative analysis of skin cancers, which is yet challenging even for dermatologists due to the inherent issues, i.e., considerable size, shape and color variation, and ambiguous boundaries. Recent vision transformers have shown promising performance in handling the variation through global context modeling. Still, they have not thoroughly solved the problem of ambiguous boundaries as they ignore the complementary usage of the boundary knowledge and global contexts. In this paper, we propose a novel cross-scale boundary-aware transformer, XBound-Former, to simultaneously address the variation and boundary problems of skin lesion segmentation. XBound-Former is a purely attention-based network and catches boundary knowledge via three specially designed learners. First, we propose an implicit boundary learner (im-Bound) to constrain the network attention on the points with noticeable boundary variation, enhancing the local context modeling while maintaining the global context. Second, we propose an explicit boundary learner (ex-Bound) to extract the boundary knowledge at multiple scales and convert it into embeddings explicitly. Third, based on the learned multi-scale boundary embeddings, we propose a cross-scale boundary learner (X-Bound) to simultaneously address the problem of ambiguous and multi-scale boundaries by using learned boundary embedding from one scale to guide the boundary-aware attention on the other scales. We evaluate the model on two skin lesion datasets and one polyp lesion dataset, where our model consistently outperforms other convolution- and transformer-based models, especially on the boundary-wise metrics. All resources could be found in https://github.com/jcwang123/xboundformer.

We present a novel deep network (namely BUSSeg) equipped with both within- and cross-image long-range dependency modeling for automated lesions segmentation from breast ultrasound images, which is a quite daunting task due to (1) the large variation of breast lesions, (2) the ambiguous lesion boundaries, and (3) the existence of speckle noise and artifacts in ultrasound images. Our work is motivated by the fact that most existing methods only focus on modeling the within-image dependencies while neglecting the cross-image dependencies, which are essential for this task under limited training data and noise. We first propose a novel cross-image dependency module (CDM) with a cross-image contextual modeling scheme and a cross-image dependency loss (CDL) to capture more consistent feature expression and alleviate noise interference. Compared with existing cross-image methods, the proposed CDM has two merits. First, we utilize more complete spatial features instead of commonly used discrete pixel vectors to capture the semantic dependencies between images, mitigating the negative effects of speckle noise and making the acquired features more representative. Second, the proposed CDM includes both intra- and inter-class contextual modeling rather than just extracting homogeneous contextual dependencies. Furthermore, we develop a parallel bi-encoder architecture (PBA) to tame a Transformer and a convolutional neural network to enhance BUSSeg's capability in capturing within-image long-range dependencies and hence offer richer features for CDM. We conducted extensive experiments on two representative public breast ultrasound datasets, and the results demonstrate that the proposed BUSSeg consistently outperforms state-of-the-art approaches in most metrics.

Detecting cells in blood smear images is of great significance for automatic diagnosis of blood diseases. However, this task is rather challenging, mainly because there are dense cells that are often overlapping, making some of the occluded boundary parts invisible. In this paper, we propose a generic and effective detection framework that exploits non-overlapping regions (NOR) for providing discriminative and confident information to compensate the intensity deficiency. In particular, we propose a feature masking (FM) to exploit the NOR mask generated from the original annotation information, which can guide the network to extract NOR features as supplementary information. Furthermore, we exploit NOR features to directly predict the NOR bounding boxes (NOR BBoxes). NOR BBoxes are combined with the original BBoxes for generating one-to-one corresponding BBox-pairs that are used for further improving the detection performance. Different from the non-maximum suppression (NMS), our proposed non-overlapping regions NMS (NOR-NMS) uses the NOR BBoxes in the BBox-pairs to calculate intersection over union (IoU) for suppressing redundant BBoxes, and consequently retains the corresponding original BBoxes, circumventing the dilemma of NMS. We conducted extensive experiments on two publicly available datasets, with positive results demonstrating the effectiveness of the proposed method against existing methods.

Deep learning models have achieved remarkable success in multi-type nuclei segmentation. These models are mostly trained at once with the full annotation of all types of nuclei available, while lack the ability of continually learning new classes due to the problem of catastrophic forgetting. In this paper, we study the practical and important class-incremental continual learning problem, where the model is incrementally updated to new classes without accessing to previous data. We propose a novel continual nuclei segmentation method, to avoid forgetting knowledge of old classes and facilitate the learning of new classes, by achieving feature-level knowledge distillation with prototype-wise relation distillation and contrastive learning. Concretely, prototype-wise relation distillation imposes constraints on the inter-class relation similarity, encouraging the encoder to extract similar class distribution for old classes in the feature space. Prototype-wise contrastive learning with a hard sampling strategy enhances the intra-class compactness and inter-class separability of features, improving the performance on both old and new classes. Experiments on two multi-type nuclei segmentation benchmarks, i.e., MoNuSAC and CoNSeP, demonstrate the effectiveness of our method with superior performance over many competitive methods. Codes are available at https://github.com/zzw-szu/CoNuSeg.

Most state-of-the-art deep networks proposed for biomedical image segmentation are developed based on U-Net. While remarkable success has been achieved, its inherent limitations hinder it from yielding more precise segmentation. First, its receptive field is limited due to the fixed kernel size, which prevents the network from modeling global context information. Second, when spatial information captured by shallower layer is directly transmitted to higher layers by skip connections, the process inevitably introduces noise and irrelevant information to feature maps and blurs their semantic meanings. In this article, we propose a novel segmentation network equipped with a new context prior guidance (CPG) module to overcome these limitations for biomedical image segmentation, namely context prior guidance network (CPG-Net). Specifically, we first extract a set of context priors under the supervision of a coarse segmentation and then employ these context priors to model the global context information and bridge the spatial-semantic gap between high-level features and low-level features. The CPG module contains two major components: context prior representation (CPR) and semantic complement flow (SCF). CPR is used to extract pixels belonging to the same objects and hence produce more discriminative features to distinguish different objects. We further introduce deep semantic information for each CPR by the SCF mechanism to compensate the semantic information diluted during the decoding. We extensively evaluate the proposed CPG-Net on three famous biomedical image segmentation tasks with diverse imaging modalities and semantic environments. Experimental results demonstrate the effectiveness of our network, consistently outperforming state-of-the-art segmentation networks in all the three tasks. Codes are available at https://github.com/zzw-szu/CPGNet .

Current representation learning methods for whole slide image (WSI) with pyramidal resolutions are inherently homogeneous and flat, which cannot fully exploit the multiscale and heterogeneous diagnostic information of different structures for comprehensive analysis. This paper presents a novel graph neural network-based multiple instance learning framework (i.e., H^2-MIL) to learn hierarchical representation from a heterogeneous graph with different resolutions for WSI analysis. A heterogeneous graph with the “resolution” attribute is constructed to explicitly model the feature and spatial-scaling relationship of multi-resolution patches. We then design a novel resolution-aware attention convolution (RAConv) block to learn compact yet discriminative representation from the graph, which tackles the heterogeneity of node neighbors with different resolutions and yields more reliable message passing. More importantly, to explore the task-related structured information of WSI pyramid, we elaborately design a novel iterative hierarchical pooling (IHPool) module to progressively aggregate the heterogeneous graph based on scaling relationships of different nodes. We evaluated our method on two public WSI datasets from the TCGA project, i.e., esophageal cancer and kidney cancer. Experimental results show that our method clearly outperforms the state-of-the-art methods on both tumor typing and staging tasks.

Automatic delineation of organ-at-risk (OAR) and gross-tumor-volume (GTV) is of great significance for radiotherapy planning. However, it is a challenging task to learn powerful representations for accurate delineation under limited pixel (voxel)-wise annotations. Contrastive learning at pixel-level can alleviate the dependency on annotations by learning dense representations from unlabeled data. Recent studies in this direction design various contrastive losses on the feature maps, to yield discriminative features for each pixel in the map. However, pixels in the same map inevitably share semantics to be closer than they actually are, which may affect the discrimination of pixels in the same map and lead to the unfair comparison to pixels in other maps. To address these issues, we propose a separated region-level contrastive learning scheme, namely SepaReg, the core of which is to separate each image into regions and encode each region separately. Specifically, SepaReg comprises two components: a structure-aware image separation (SIS) module and an intra- and inter-organ distillation (IID) module. The SIS is proposed to operate on the image set to rebuild a region set under the guidance of structural information. The inter-organ representation will be learned from this set via typical contrastive losses cross regions. On the other hand, the IID is proposed to tackle the quantity imbalance in the region set as tiny organs may produce fewer regions, by exploiting intra-organ representations. We conducted extensive experiments to evaluate the proposed model on a public dataset and two private datasets. The experimental results demonstrate the effectiveness of the proposed model, consistently achieving better performance than state-of-the-art approaches. Code is available at https://github.com/jcwang123/Separate_CL.

Can you find me? By simulating how humans to discover the so-called 'perfectly'-camouflaged object, we present a novel boundary-guided separated attention network (call BSA-Net). Beyond the existing camouflaged object detection (COD) wisdom, BSA-Net utilizes two-stream separated attention modules to highlight the separator (or say the camouflaged object's boundary) between an image's background and foreground: the reverse attention stream helps erase the camouflaged object's interior to focus on the background, while the normal attention stream recovers the interior and thus pay more attention to the foreground; and both streams are followed by a boundary guider module and combined to strengthen the understanding of boundary. The core design of such separated attention is motivated by the COD procedure of humans: find the subtle difference between the foreground and background to delineate the boundary of a camouflaged object, then the boundary can help further enhance the COD accuracy. We validate on three benchmark datasets that the proposed BSA-Net is very beneficial to detect camouflaged objects with the blurred boundaries and similar colors/patterns with their backgrounds. Extensive results exhibit very clear COD improvements on our BSA-Net over sixteen SOTAs.

Magnetic resonance imaging (MRI) can present multicontrast images of the same anatomical structures, enabling multi-contrast super-resolution (SR) techniques. Compared with SR reconstruction using a single-contrast, multicontrast SR reconstruction is promising to yield SR images with higher quality by leveraging diverse yet complementary information embedded in different imaging modalities. However, existing methods still have two shortcomings: (1) they neglect that the multi-contrast features at different scales contain different anatomical details and hence lack effective mechanisms to match and fuse these features for better reconstruction; and (2) they are still deficient in capturing long-range dependencies, which are essential for the regions with complicated anatomical structures. We propose a novel network to comprehensively address these problems by developing a set of innovative Transformer-empowered multi-scale contextual matching and aggregation techniques; we call it McMRSR. Firstly, we tame transformers to model long-range dependencies in both reference and target images. Then, a new multi-scale contextual matching method is proposed to capture corresponding contexts from reference features at different scales. Furthermore, we introduce a multi-scale aggregation mechanism to gradually and interactively aggregate multi-scale matched features for reconstructing the target SR MR image. Extensive experiments demonstrate that our network outperforms state-of-the-art approaches and has great potential to be applied in clinical practice. Codes are available at https://github.com/XAIMI-Lab/McMRSR.

We study the semi-supervised learning problem, using a few labeled data and a large amount of unlabeled data to train the network, by developing a cross-patch dense contrastive learning framework, to segment cellular nuclei in histopathologic images. This task is motivated by the expensive burden on collecting labeled data for histopathologic image segmentation tasks. The key idea of our method is to align features of teacher and student networks, sampled from cross-image in both patch- and pixel-levels, for enforcing the intra-class compactness and inter-class separability of features that as we shown is helpful for extracting valuable knowledge from unlabeled data. We also design a novel optimization framework that combines consistency regularization and entropy minimization techniques, showing good property in eviction of gradient vanishing. We assess the proposed method on two publicly available datasets, and obtain positive results on extensive experiments, outperforming the state-of-the-art methods. Codes are available at https://github.com/zzw-szu/CDCL.

Rain severely degrades the visibility of scene objects, especially when images are captured through the glass under rainy weather. We observe three intriguing phenomena: 1) rain is a mixture of raindrops, rain streaks and rainy haze; 2) the depth from the camera determines the degree of object visibility, where objects nearby and far away are visually blocked by rain streaks and rainy haze, respectively; and 3) raindrops on the glass randomly affect the object visibility of the whole image space. However, existing solutions and benchmark datasets lack full consideration of the mixture of rain (MOR). In this paper, we originally consider that the overall object visibility is determined by MOR, and enrich the RainCityscapes by considering real-world raindrops to construct the MOR dataset, named RainCityscapes++. To solve the practical rain removal problem arisen from MOR, we formulate a new rain imaging model and propose a multi-branch attention generative adversarial network (MBA-RainGAN). Extensive experiments show clear improvements of our approach over SOTAs on RainCityscapes++.

Conventional methods address the cross-modal retrieval problem by projecting the multi-modal data into a shared representation space. Such a strategy will inevitably lose the modality-specific information, leading to decreased retrieval accuracy. In this paper, we propose heterogeneous graph embeddings to preserve more abundant cross-modal information. The embedding from one modality will be compensated with the aggregated embeddings from the other modality. In particular, a self-denoising tree search is designed to reduce the "label noise" problem, making the heterogeneous neighborhood more semantically relevant. The dual-path aggregation tackles the "modality imbalance" problem, giving each sample comprehensive dual-modality information. The final heterogeneous graph embedding is obtained by feeding the aggregated dual-modality features to the cross-modal self-attention module. Experiments conducted on cross-modality person re-identification and image-text retrieval task validate the superiority and generality of the proposed method.

Video dehazing using deep-learning based methods has just received increasing attention in recent years. However, most existing methods tackle temporal consistency in the color domain only, which are less sensitive to small and imperceptible motions in a video, due to fog's drift and diffusion. In this work, we investigate in the frequency domain, which enables us to capture small motions effectively, and find that the phase component contains more semantic structures yet less haze information than the amplitude component of the hazy image. Based on these observations, we propose a novel phase-based memory network (PM-Net) to integrate the phase and color memory information for boosting video dehazing. Apart from the color memory from consecutive video frames, our PM-Net constructs a phase memory, which stores phase features of past video frames, and devise a cross-modal memory read (CMR) module, which fully leverages features from the color memory and the phase memory to boost features extracted from the current video frame for dehazing. Experimental results on the benchmark dataset of real hazy videos and a newly collected dataset of synthetic videos, show that the proposed PM-Net clearly outperforms the state-of-the-art image and video dehazing methods. Code is available at https://github.com/liuye123321/PM-Net.

Abstract High‐confidence overlap prediction and accurate correspondences are critical for cutting‐edge models to align paired point clouds in a partial‐to‐partial manner. However, there inherently exists uncertainty between the overlapping and non‐overlapping regions, which has always been neglected and significantly affects the registration performance. Beyond the current wisdom, we propose a novel uncertainty‐aware overlap prediction network, dubbed UTOPIC, to tackle the ambiguous overlap prediction problem; to our knowledge, this is the first to explicitly introduce overlap uncertainty to point cloud registration. Moreover, we induce the feature extractor to implicitly perceive the shape knowledge through a completion decoder, and present a geometric relation embedding for Transformer to obtain transformation‐invariant geometry‐aware feature representations. With the merits of more reliable overlap scores and more precise dense correspondences, UTOPIC can achieve stable and accurate registration results, even for the inputs with limited overlapping areas. Extensive quantitative and qualitative experiments on synthetic and real benchmarks demonstrate the superiority of our approach over state‐of‐the‐art methods.

As a novel similarity measure that is defined as the expectation of a kernel function between two random variables, correntropy has been successfully applied in robust machine learning and signal processing to combat large outliers. The kernel function in correntropy is usually a zero-mean Gaussian kernel. In a recent work, the concept of mixture correntropy (MC) was proposed to improve the learning performance, where the kernel function is a mixture Gaussian kernel, namely, a linear combination of several zero-mean Gaussian kernels with different widths. In both correntropy and MC, the center of the kernel function is, however, always located at zero. In the present work, to further improve the learning performance, we propose the concept of multikernel correntropy (MKC), in which each component of the mixture Gaussian kernel can be centered at a different location. The properties of the MKC are investigated and an efficient approach is proposed to determine the free parameters in MKC. Experimental results show that the learning algorithms under the maximum MKC criterion (MMKCC) can outperform those under the original maximum correntropy criterion (MCC) and the maximum MC criterion (MMCC).

Camouflaged objects share very similar colors but have different semantics with the surroundings. Cognitive scientists observe that both the global contour (i.e., boundary) and the local pattern (i.e., texture) of camouflaged objects are key cues to help humans find them successfully. Inspired by the cognitive scientist's observation, we propose a novel boundary-and-texture enhancement network (FindNet) for camouflaged object detection (COD) from single images. Different from most of existing COD methods, FindNet embeds both the boundary-and-texture information into the camouflaged object features. The boundary enhancement (BE) module is leveraged to focus on the global contour of the camouflaged object, and the texture enhancement (TE) module is utilized to focus on the local pattern. The enhanced features from BE and TE, which complement each other, are combined to obtain the final prediction. FindNet performs competently on various conditions of COD, including slightly clear boundaries but very similar textures, fuzzy boundaries but slightly differentiated textures, and simultaneous fuzzy boundaries and textures. Experimental results exhibit clear improvements of FindNet over fifteen state-of-the-art methods on four benchmark datasets, in terms of detection accuracy and boundary clearness. The code will be publicly released.

Automatic vertebra segmentation from computed tomography (CT) image is the very first and a decisive stage in vertebra analysis for computer-based spinal diagnosis and therapy support system. However, automatic segmentation of vertebra remains challenging due to several reasons, including anatomic complexity of spine, unclear boundaries of the vertebrae associated with spongy and soft bones. Based on 2D U-Net, we have proposed an Embedded Clustering Sliced U-Net (ECSU-Net). ECSU-Net comprises of three modules named segmentation, intervertebral disc extraction (IDE) and fusion. The segmentation module follows an instance embedding clustering approach, where our three sliced sub-nets use axis of CT images to generate a coarse 2D segmentation along with embedding space with the same size of the input slices. Our IDE module is designed to classify vertebra and find the inter-space between two slices of segmented spine. Our fusion module takes the coarse segmentation (2D) and outputs the refined 3D results of vertebra. A novel adaptive discriminative loss (ADL) function is introduced to train the embedding space for clustering. In the fusion strategy, three modules are integrated via a learnable weight control component, which adaptively sets their contribution. We have evaluated classical and deep learning methods on Spineweb dataset-2. ECSU-Net has provided comparable performance to previous neural network based algorithms achieving the best segmentation dice score of 95.60% and classification accuracy of 96.20%, while taking less time and computation resources.

Parkinson’s disease (PD) is known as an irreversible neurodegenerative disease that mainly affects the patient’s motor system. Early classification and regression of PD are essential to slow down this degenerative process from its onset. In this article, a novel adaptive unsupervised feature selection approach is proposed by exploiting manifold learning from longitudinal multimodal data. Classification and clinical score prediction are performed jointly to facilitate early PD diagnosis. Specifically, the proposed approach performs united embedding and sparse regression, which can determine the similarity matrices and discriminative features adaptively. Meanwhile, we constrain the similarity matrix among subjects and exploit the${l}_{\mathrm {2,p}}$norm to conduct sparse adaptive control for obtaining the intrinsic information of the multimodal data structure. An effective iterative optimization algorithm is proposed to solve this problem. We perform abundant experiments on the Parkinson’s Progression Markers Initiative (PPMI) data set to verify the validity of the proposed approach. The results show that our approach boosts the performance on the classification and clinical score regression of longitudinal data and surpasses the state-of-the-art approaches.

We present a novel deep learning model equipped with a new region-aware global context modeling technique for automatic nerve segmentation from ultrasound images, which is a challenging task due to (1) the large variation and blurred boundaries of targets, (2) the large amount of speckle noise in ultrasound images, and (3) the inherent real-time requirement of this task. It is essential to efficiently capture long-range dependencies by global context modeling for a segmentation network to overcome these challenges. Traditional global context modeling techniques usually explore pixel-aware correlations to establish long-range dependencies, which are usually computation-intensive and greatly degrade time performance. In addition, in this application, pixel-aware modeling may inevitably introduce much speckle noise in the computation and potentially degrade segmentation performance. In this paper, we propose a novel region-aware modeling technique to establish long-range dependencies based on different regions to improve segmentation accuracy while maintaining real-time performance; we call it region-aware pyramid aggregation (RPA) module. In order to adaptively divide the feature maps into a set of semantic-independent regions, we develop an attention mechanism and integrate it into the spatial pyramid network to evaluate the semantic similarity of different regions. We further develop an adaptive pyramid fusion (APF) module to dynamically fuse the multi-level features generated from the decoder to refining the segmentation results. We conducted extensive experiments on a famous public ultrasound nerve image segmentation dataset. Experimental results demonstrate that our method consistently outperforms our rivals in terms of segmentation accuracy. The code is available at https://github.com/jsonliu-szu/RAGCM.

We propose a novel convolutional neural network (ConvNet) equipped with two new semantic calibration and refinement approaches for automatic polyp segmentation from colonoscopy videos. While ConvNets set state-of-the-are performance for this task, it is still difficult to achieve satisfactory results in a real-time manner, which is a necessity in clinical practice. The main obstacle is the huge semantic gap between high-level features and low-level features, making it difficult to take full advantage of complementary semantic information contained in these hierarchical features. Compared with existing solutions, which either directly aggregate these features without considering the semantic gap or employ sophisticated non-local modeling techniques to refine semantic information by introduce many extra computational costs, the proposed ConvNet is able to more precisely yet efficiently calibrate and refine semantic information for better segmentation performance without increasing model complexity; we call the proposed ConvNet as SCR-Net, which has two key modules. We first propose a semantic calibration module (SCM) to effectively transmit the semantic information from high-level layers to low-level layers by learning the semantic-spatial relations during the training procedure. We then propose a semantic refinement module (SRM) to, based on the features calibrated by SCM, enhance the discrimination capability of the features for targeting objects. Extensive experiments on the Kvasir-SEG dataset demonstrate that the proposed SCR-Net is capable of achieving better segmentation accuracy than state-of-the-art approaches with a faster speed. The proposed techniques are general enough to be applied to similar applications where precise and efficient multi-level feature fusion is critical. The code is available at https://github.com/jiafuz/SCR-Net.

Shadow detection in a single image has received significant research interests in recent years. However, much fewer works have been explored in shadow detection over dynamic scenes. The bottleneck is the lack of a well-established dataset with high-quality annotations for video shadow detection. In this work, we collect a new video shadow detection dataset (ViSha), which contains 120 videos with 11,685 frames, covering 60 object categories, varying lengths, and different motion/lighting conditions. All the frames are annotated with a high-quality pixel-level shadow mask. To the best of our knowledge, this is the first learning-oriented dataset for video shadow detection. Furthermore, we develop a new baseline model, named triple-cooperative video shadow detection network (TVSD-Net). It utilizes triple parallel networks in a cooperative manner to learn discriminative representations at intra-video and inter-video levels. Within the network, a dual gated co-attention module is proposed to constrain features from neighboring frames in the same video, while an auxiliary similarity loss is introduced to mine semantic information between different videos. Finally, we conduct a comprehensive study on ViSha, evaluating 12 state-of-the-art models (including single image shadow detectors, video object segmentation, and saliency detection methods). Experiments demonstrate that our model outperforms SOTA competitors.

Federated learning allows distributed medical institutions to collaboratively learn a shared prediction model with privacy protection. While at clinical deployment, the models trained in federated learning can still suffer from performance drop when applied to completely unseen hospitals outside the federation. In this paper, we point out and solve a novel problem setting of federated domain generalization (FedDG), which aims to learn a federated model from multiple distributed source domains such that it can directly generalize to unseen target domains. We present a novel approach, named as Episodic Learning in Continuous Frequency Space (ELCFS), for this problem by enabling each client to exploit multi-source data distributions under the challenging constraint of data decentralization. Our approach transmits the distribution information across clients in a privacy-protecting way through an effective continuous frequency space interpolation mechanism. With the transferred multi-source distributions, we further carefully design a boundary-oriented episodic learning paradigm to expose the local learning to domain distribution shifts and particularly meet the challenges of model generalization in medical image segmentation scenario. The effectiveness of our method is demonstrated with superior performance over state-of-the-arts and in-depth ablation experiments on two medical image segmentation tasks. The code is available at https://github.com/liuquande/FedDG-ELCFS.

Automatic polyp segmentation from colonoscopy images is an essential step in computer aided diagnosis for colorectal cancer. Most of polyp segmentation methods reported in recent years are based on fully supervised deep learning. However, annotation for polyp images by physicians during the diagnosis is time-consuming and costly. In this paper, we present a novel semi-supervised polyp segmentation via collaborative and adversarial learning of focused and dispersive representations learning model, where focused and dispersive extraction module are used to deal with the diversity of location and shape of polyps. In addition, confidence maps produced by a discriminator in an adversarial training framework shows the effectiveness of leveraging unlabeled data and improving the performance of segmentation network. Consistent regularization is further employed to optimize the segmentation networks to strengthen the representation of the outputs of focused and dispersive extraction module. We also propose an auxiliary adversarial learning method to better leverage unlabeled examples to further improve semantic segmentation accuracy. We conduct extensive experiments on two famous polyp datasets: Kvasir-SEG and CVC-Clinic DB. Experimental results demonstrate the effectiveness of the proposed model, consistently outperforming state-of-the-art semi-supervised segmentation models based on adversarial training and even some advanced fully supervised models.

Adaptive person re-identification (adaptive ReID) targets at transferring learned knowledge from the labeled source domain to the unlabeled target domain. Pseudo-label-based methods that alternatively generate pseudo labels and optimize the training model have demonstrated great effectiveness in this field. However, the generated pseudo labels are inaccurate and cannot reflect the true semantic meaning of the unlabeled samples. We consider such inaccuracy stems from both the lagged update of the pseudo labels as well as the simple criterion of the employed clustering method. To tackle the problem, we propose an online pseudo label generation by hierarchical cluster dynamics for adaptive ReID. In particular, hierarchical label banks are constructed for all the samples in the dataset, and we update the pseudo labels of the sample in each coming mini-batch, performing the model optimization and the label generation simultaneously. A new hierarchical cluster dynamics is built for the label update, where cluster merge and cluster split are driven by a possibility computed by the label propagation. Our method can achieve better pseudo labels and higher reid accuracy. Extensive experiments on Market-to-Duke, Duke-to-Market, MSMT-to-Market, MSMT-to-Duke, Market-to-MSMT, and Duke-to-MSMT verify the effectiveness of our proposed method.

Convolution on 3D point clouds that generalized from 2D grid-like domains is widely researched yet far from perfect. The standard convolution characterises feature correspondences indistinguishably among 3D points, presenting an intrinsic limitation of poor distinctive feature learning. In this paper, we propose Adaptive Graph Convolution (AdaptConv) which generates adaptive kernels for points according to their dynamically learned features. Compared with using a fixed/isotropic kernel, AdaptConv improves the flexibility of point cloud convolutions, effectively and precisely capturing the diverse relations between points from different semantic parts. Unlike popular attentional weight schemes, the proposed AdaptConv implements the adaptiveness inside the convolution operation instead of simply assigning different weights to the neighboring points. Extensive qualitative and quantitative evaluations show that our method outperforms state-of-the-art point cloud classification and segmentation approaches on several benchmark datasets. Our code is available at https://github.com/hrzhou2/AdaptConv-master.

We investigate vehicle density detection from traffic surveillance. This task is rather challenging, mainly due to the low-resolution of data and large-scale variance of vehicles. The main result is that by learning cross-scale features, high-quality vehicle density maps can be attainable. Our main technical contribution is a learning model, called Adaptive Cross-Scale Network (ACSNet), that can learn cross-scale features from traffic surveillance data with low-resolution and large scale variance of vehicles. ACSNet consists of 1) a series of cross-scale feature extraction blocks with dense bypassing paths for harvesting spatial information, 2) an attention block for learning from appropriate scales, and 3) a structural similarity index for learning from occlusion scenes. We assess our ACSNet on two benchmark datasets, and extensive empirical evidence shows that our ACSNet performs favor-ably against the state-of-the-art methods.

We present a novel direction-aware feature-level frequency decomposition network for single image deraining. Compared with existing solutions, the proposed network has three compelling characteristics. First, unlike previous algorithms, we propose to perform frequency decomposition at feature-level instead of image-level, allowing both low-frequency maps containing structures and high-frequency maps containing details to be continuously refined during the training procedure. Second, we further establish communication channels between low-frequency maps and high-frequency maps to interactively capture structures from high-frequency maps and add them back to low-frequency maps and, simultaneously, extract details from low-frequency maps and send them back to high-frequency maps, thereby removing rain streaks while preserving more delicate features in the input image. Third, different from existing algorithms using convolutional filters consistent in all directions, we propose a direction-aware filter to capture the direction of rain streaks in order to more effectively and thoroughly purge the input images of rain streaks. We extensively evaluate the proposed approach in three representative datasets and experimental results corroborate our approach consistently outperforms state-of-the-art deraining algorithms.

Deep neural networks have been shown to be very powerful tools for object detection in various scenes. Their remarkable performance, however, heavily depends on the availability of a large number of high quality labeled data, which are time-consuming and costly to acquire for scenes with densely packed objects. We present a novel semi-supervised approach to addressing this problem, which is designed based on a common teacher-student model, integrated with a novel intersection-over-union (IoU) aware consistency loss and a new proposal consistency loss. The IoU-aware consistency loss evaluates the IoU over the prediction pairs of the teacher model and the student model, which enforces the prediction of the student model to approach closely to that of the teacher model. The IoU-aware consistency loss also reweights the importance of different prediction pairs to suppress the low-confident pairs. The proposal consistency loss ensures proposal consistency between the two models, making it possible to involve the region proposal network in the training process with unlabeled data. We also construct a new dataset, namely RebarDSC, containing 2,125 rebar images annotated with 350,348 bounding boxes in total (164.9 annotations per image average), to evaluate the proposed method. Extensive experiments are conducted over both the RebarDSC dataset and the famous large public dataset SKU-110K. Experimental results corroborate that the proposed method is able to improve the object detection performance in densely packed scenes, consistently outperforming state-of-the-art approaches. Dataset is available in https://github.com/Armin1337/RebarDSC.

Automated surgical gesture recognition is of great importance in robot-assisted minimally invasive surgery. However, existing methods assume that training and testing data are from the same domain, which suffers from severe performance degradation when a domain gap exists, such as the simulator and real robot. In this paper, we propose a novel unsupervised domain adaptation framework which can simultaneously transfer multi-modality knowledge, i.e., both kinematic and visual data, from simulator to real robot. It remedies the domain gap with enhanced transferable features by using temporal cues in videos, and inherent correlations in multi-modal towards recognizing gesture. Specifically, we first propose a Motion Direction Oriented Kinematics feature alignment (MDO-K) to align kinematics, which exploits temporal continuity to transfer motion directions with smaller gap rather than position values, relieving the adaptation burden. Moreover, we propose a Kinematic and Visual Relation Attention (KV-Relation-ATT) to transfer the co-occurrence signals of kinematics and vision. Such features attended by correlation similarity are more informative for enhancing domain-irreverent of the model. Two feature alignment strategies benefit the model mutually during the end-to-end learning process. We extensively evaluate our method for gesture recognition using DESK dataset with peg transfer procedure. Results show that our approach recovers the performance with great improvement gains, up to 12.91% in Accuracy and 20.16% in F1score without using any annotations in real robot.

Single image dehazing is a challenging task, for which the domain shift between synthetic training data and real-world testing images usually leads to degradation of existing methods. To address this issue, we propose a novel image dehazing framework collaborating with unlabeled real data. First, we develop a disentangled image dehazing network (DID-Net), which disentangles the feature representations into three component maps, i.e. the latent haze-free image, the transmission map, and the global atmospheric light estimate, respecting the physical model of a haze process. Our DID-Net predicts the three component maps by progressively integrating features across scales, and refines each map by passing an independent refinement network. Then a disentangled-consistency mean-teacher network (DMT-Net) is employed to collaborate unlabeled real data for boosting single image dehazing. Specifically, we encourage the coarse predictions and refinements of each disentangled component to be consistent between the student and teacher networks by using a consistency loss on unlabeled real data. We make comparison with 13 state-of-the-art dehazing methods on a new collected dataset (Haze4K) and two widely-used dehazing datasets (i.e., SOTS and HazeRD), as well as on real-world hazy images. Experimental results demonstrate that our method has obvious quantitative and qualitative improvements over the existing methods.

Verifiable pattern matching enables users to obtain authenticated query results over outsourced data on an untrusted remote server. It is a fundamental problem in many security-critical big data applications, including big database search, human genome data search, text search, etc., especially when these applications are outsourced to third-party clouds. However, the state-of-the-art schemes do not yet support efficient data updates. In this work, we propose the first dynamic verifiable pattern matching scheme to support efficient data updates. The proposed scheme is built on two ideas: one is to embed unique randomness to decouple the character and its index in the outsourced data, enabling efficient data updates; the other is to reduce the verifiable pattern matching problem to a discrete set membership testing problem, which relies on the decoupling introduced in the first idea. Based on these two ideas, the proposed scheme first employs the suffix array index structure to search pattern matching queries. The scheme then authenticates the outsourced text using a newly designed authenticated data structure based on the RSA accumulator, which guarantees the verifiability of pattern matching query results. Data update is naturally supported using the RSA accumulator working on discrete sets. Based on the proposed design, we have prototyped a proof-of-concept for the proposed scheme and have conducted an extensive experimental evaluation. In addition to supporting efficient data update, our experimental results show that the proposed scheme incurs reduced verification cost in comparison with the baseline state-of-the-art scheme.

Tumor metastases detection is of great importance for the treatment of breast cancer patients. Various CNN (convolutional neural network) based methods get excellent performance in object detection/segmentation. However, the detection of metastases in hematoxylin and eosin (H&E) stained whole-slide images (WSI) is still challenging mainly due to two aspects. (1) The resolution of the image is too large. (2) lacking labeled training data. Whole-slide images generally stored in a multi-resolution structure with multiple downsampled tiles. It is difficult to feed the whole image into memory without compression. Moreover, labeling images for the pathologists are time-consuming and expensive. In this paper, we study the problem of detecting breast cancer metastases in the pathological image on patch level. To address the abovementioned challenges, we propose a few-shot learning method to classify whether an image patch contains tumor cells. Specifically, we propose a patch-level unsupervised cell ranking approach, which only relies on images with limited labels. The main idea of the proposed method is that when cropping a patch A from the WSI and further cropping a sub-patch B from A, the cell number of A is always larger than that of B. Based on this observation, we make use of the unlabeled images to learn the ranking information of cell counting to extract the abstract features. Experimental results show that our method is effective to improve the patch-level classification accuracy, compared to the traditional supervised method. The source code is publicly available at https://github.com/fewshot-camelyon.

This work presents a gated non-local deep residual learning framework for image deraining. It can avoid the over-deraining or under-deraining caused by the global residual learning in existing deraining networks, since the learned soft gate in our method adaptively adjusts the amount of global residual to be passed for generating the final derained result. To generate feature maps for global residual prediction, we develop a non-local guided attention module (NLAM), which first obtains non-local features by exploiting spatial inter-dependencies among all the feature positions of local features produced by convolutional neural network (CNN), and then leverages the attention mechanism to merge the local and non-local features based on their complementary relation. Moreover, we develop a channel-wise gated prediction module to learn a soft gate on the global residual by explicitly modelling channel inter-dependencies of the feature maps obtained from NLAM. Experiments on four deraining benchmark datasets and real-world rainy images show that our network has a quantitative and qualitative improvement over state-of-the-arts.

This article substantially advances upon state-of-the-art to enhance liver vessels segmentation accuracy by leveraging advantages of synthetic PET-CT (SPET-CT) images in addition to computed tomography angiography (CTA) volumes. Our setup makes a hybrid solution of modified generative adversarial network-convolutional autoencoder (GAN-cAED) combining synthetic ability of GAN to deliver SPET-CT images with generative ability of cAED network in terms of latent learning to more refined segmentation of major liver vessels. We improve time complexity through a novel concept of controlled segmentation by introducing a threshold metric to stop segmentation up to a desired level. The innovative concept of controlled vessel segmentation with a stopping criterion via variant threshold levels will help surgeons to avoid unintentional major blood vessels cutting, reducing the risk of excessive blood loss. Clinically, such solutions offer computer-aided liver surgeries and drug treatment evaluation in a CTA-only environment, shorten the requirement of radioactive and expensive fused PET-CT images.

Recently, an emerging trend in medical image classification is to combine radiomics framework with deep learning classification network in an integrated system. Although this combination is efficient in some tasks, the deep learning-based classification network is often difficult to capture an effective representation of lesion regions, and prone to face the challenge of overfitting, leading to unreliable features and inaccurate results, especially when the sizes of the lesions are small or the training dataset is small. In addition, these combinations mostly lack an effective feature selection mechanism, which makes it difficult to obtain the optimal feature selection. In this paper, we introduce a novel and effective deep semantic segmentation feature-based radiomics (DSFR) framework to overcome the above-mentioned challenges, which consists of two modules: the deep semantic feature extraction module and the feature selection module. Specifically, the extraction module is utilized to extract hierarchical semantic features of the lesions from a trained segmentation network. The feature selection module aims to select the most representative features by using a novel feature similarity adaptation algorithm. Experiments are extensively conducted to evaluate our method in two clinical tasks: the pathological grading prediction in pancreatic neuroendocrine neoplasms (pNENs), and the prediction of thrombolytic therapy efficacy in deep venous thrombosis (DVT). Experimental results on both tasks demonstrate that the proposed method consistently outperforms the state-of-the-art approaches by a large margin.

Restoring the normal masticatory function of broken teeth is a challenging task primarily due to the defect location and size of a patient's teeth. In recent years, although some representative image-to-image transformation methods (e.g. Pix2Pix) can be potentially applicable to restore the missing crown surface, most of them fail to generate dental inlay surface with realistic crown details (e.g. occlusal groove) that are critical to the restoration of defective teeth with varying shapes. In this article, we design a computer-aided Deep Adversarial-driven dental Inlay reStoration (DAIS) framework to automatically reconstruct a realistic surface for a defective tooth. Specifically, DAIS consists of a Wasserstein generative adversarial network (WGAN) with a specially designed loss measurement, and a new local-global discriminator mechanism. The local discriminator focuses on missing regions to ensure the local consistency of a generated occlusal surface, while the global discriminator aims at defective teeth and adjacent teeth to assess if it is coherent as a whole. Experimental results demonstrate that DAIS is highly efficient to deal with a large area of missing teeth in arbitrary shapes and generate realistic occlusal surface completion. Moreover, the designed watertight inlay prostheses have enough anatomical morphology, thus providing higher clinical applicability compared with more state-of-the-art methods.

We present a convolutional neural network (CNN) equipped with a novel and efficient adaptive dual attention module (ADAM) for automated skin lesion segmentation from dermoscopic images, which is an essential yet challenging step for the development of a computer-assisted skin disease diagnosis system. The proposed ADAM has three compelling characteristics. First, we integrate two global context modeling mechanisms into the ADAM, one aiming at capturing the boundary continuity of skin lesion by global average pooling while the other dealing with the shape irregularity by pixel-wise correlation. In this regard, our network, thanks to the proposed ADAM, is capable of extracting more comprehensive and discriminative features for recognizing the boundary of skin lesions. Second, the proposed ADAM supports multi-scale resolution fusion, and hence can capture multi-scale features to further improve the segmentation accuracy. Third, as we harness a spatial information weighting method in the proposed network, our method can reduce a lot of redundancies compared with traditional CNNs. The proposed network is implemented based on a dual encoder architecture, which is able to enlarge the receptive field without greatly increasing the network parameters. In addition, we assign different dilation rates to different ADAMs so that it can adaptively capture distinguishing features according to the size of a lesion. We extensively evaluate the proposed method on both ISBI2017 and ISIC2018 datasets and the experimental results demonstrate that, without using network ensemble schemes, our method is capable of achieving better segmentation performance than state-of-the-art deep learning models, particularly those equipped with attention mechanisms.

Automatic gastric tumor segmentation and lymph node (LN) classification not only can assist radiologists in reading images, but also provide image-guided clinical diagnosis and improve diagnosis accuracy. However, due to the inhomogeneous intensity distribution of gastric tumor and LN in CT scans, the ambiguous/missing boundaries, and highly variable shapes of gastric tumor, it is quite challenging to develop an automatic solution. To comprehensively address these challenges, we propose a novel 3D multi-attention guided multi-task learning network for simultaneous gastric tumor segmentation and LN classification, which makes full use of the complementary information extracted from different dimensions, scales, and tasks. Specifically, we tackle task correlation and heterogeneity with the convolutional neural network consisting of scale-aware attention-guided shared feature learning for refined and universal multi-scale features, and task-aware attention-guided feature learning for task-specific discriminative features. This shared feature learning is equipped with two types of scale-aware attention (visual attention and adaptive spatial attention) and two stage-wise deep supervision paths. The task-aware attention-guided feature learning comprises a segmentation-aware attention module and a classification-aware attention module. The proposed 3D multi-task learning network can balance all tasks by combining segmentation and classification loss functions with weight uncertainty. We evaluate our model on an in-house CT images dataset collected from three medical centers. Experimental results demonstrate that our method outperforms the state-of-the-art algorithms, and obtains promising performance for tumor segmentation and LN classification. Moreover, to explore the generalization for other segmentation tasks, we also extend the proposed network to liver tumor segmentation in CT images of the MICCAI 2017 Liver Tumor Segmentation Challenge. Our implementation is released at https://github.com/infinite-tao/MA-MTLN.

With the explosive growth of action categories, zero-shot action recognition aims to extend a well-trained model to novel/unseen classes. To bridge the large knowledge gap between seen and unseen classes, in this brief, we visually associate unseen actions with seen categories in a visually connected graph, and the knowledge is then transferred from the visual features space to semantic space via the grouped attention graph convolutional networks (GAGCNs). In particular, we extract visual features for all the actions, and a visually connected graph is built to attach seen actions to visually similar unseen categories. Moreover, the proposed grouped attention mechanism exploits the hierarchical knowledge in the graph so that the GAGCN enables propagating the visual-semantic connections from seen actions to unseen ones. We extensively evaluate the proposed method on three data sets: HMDB51, UCF101, and NTU RGB + D. Experimental results show that the GAGCN outperforms state-of-the-art methods.

In recent years, vaccine incidents occurred around the world, which endangers people’s lives. In the technical respect, these incidents are partially due to the fact that existing vaccine management systems are distributively managed by different entities in the vaccine supply chain. This architecture makes it relatively easy to modify or even delete the vaccine circulation data maliciously, which makes tracing problematic vaccine hard and identifying the responsibility for a vaccine accident hard. To solve these issues, this article presents a blockchain-based solution to protect the whole process of vaccine circulation. We first propose a model to supervise the vaccine circulation process by incorporating existing regulatory practices. Then, we propose a blockchain-based tracing system to implement this model. The proposed system takes the blockchain as a global, unique, and verifiable database to store all the circulation data. Through data insertions and queries on the global and unique database, the proposed system achieves the protection of vaccine circulation. We also implement a proof-of-concept prototype of the proposed system. Experimental results confirm that the proposed system is beneficial.

In this paper, we propose a normal estimation method for unstructured point cloud. We observe that geometric estimators commonly focus more on feature preservation but are hard to tune parameters and sensitive to noise, while learning-based approaches pursue an overall normal estimation accuracy but cannot well handle challenging regions such as surface edges. This paper presents a novel normal estimation method, under the co-support of geometric estimator and deep learning. To lowering the learning difficulty, we first propose to compute a suboptimal initial normal at each point by searching for a best fitting patch. Based on the computed normal field, we design a normal-based height map network (NH-Net) to fine-tune the suboptimal normals. Qualitative and quantitative evaluations demonstrate the clear improvements of our results over both traditional methods and learning-based methods, in terms of estimation accuracy and feature recovery.

Abstract We propose a novel deep example‐based image colourization method called dense encoding pyramid network. In our study, we define the colourization as a multinomial classification problem. Given a greyscale image and a reference image, the proposed network leverages large‐scale data and then predicts colours by analysing the colour distribution of the reference image. We design the network as a pyramid structure in order to exploit the inherent multi‐scale, pyramidal hierarchy of colour representations. Between two adjacent levels, we propose a hierarchical decoder–encoder filter to pass the colour distributions from the lower level to higher level in order to take both semantic information and fine details into account during the colourization process. Within the network, a novel parallel residual dense block is proposed to effectively extract the local–global context of the colour representations by widening the network. Several experiments, as well as a user study, are conducted to evaluate the performance of our network against state‐of‐the‐art colourization methods. Experimental results show that our network is able to generate colourful, semantically correct and visually pleasant colour images. In addition, unlike fully automatic colourization that produces fixed colour images, the reference image of our network is flexible; both natural images and simple colour palettes can be used to guide the colourization.

Shadow detection and shadow removal are fundamental and challenging tasks, requiring an understanding of the global image semantics. This paper presents a novel deep neural network design for shadow detection and removal by analyzing the spatial image context in a direction-aware manner. To achieve this, we first formulate the direction-aware attention mechanism in a spatial recurrent neural network (RNN) by introducing attention weights when aggregating spatial context features in the RNN. By learning these weights through training, we can recover direction-aware spatial context (DSC) for detecting and removing shadows. This design is developed into the DSC module and embedded in a convolutional neural network (CNN) to learn the DSC features at different levels. Moreover, we design a weighted cross entropy loss to make effective the training for shadow detection and further adopt the network for shadow removal by using a euclidean loss function and formulating a color transfer function to address the color and luminosity inconsistencies in the training pairs. We employed two shadow detection benchmark datasets and two shadow removal benchmark datasets, and performed various experiments to evaluate our method. Experimental results show that our method performs favorably against the state-of-the-art methods for both shadow detection and shadow removal.

This paper presents a novel deep learning model to aggregate the attentional dilated features for salient object detection by exploring the complementary information between the global and local context in a convolutional neural network. There are two technical contributions to our network design. First, we develop an attentional dense atrous (dilated) spatial pyramid pooling (AD-ASPP) module to selectively use the local saliency cues captured by dilated convolutions with a small rate and the global saliency cues captured by dilated convolutions with a large rate. Second, taking the feature pyramid network as the backbone, we develop an aggregation network to integrate the refined features by formulating two consecutive chains of residual learning based modules: one chain from deep to shallow layers while another chain from shallow to deep layers. We evaluate our network on seven widely-used saliency detection benchmarks by comparing it against 21 state-of-the-art methods. Experimental results show that our network outperforms others on all the seven benchmark datasets.

This article presents an efficient superblock-based sample adaptive offset (superSAO) that jointly exploits block-wise filter partition and probabilistic band interval segmentation to improve quality-of-experience of industrial video applications. Specifically, we investigate the partition flexibility of a superSAO block whose root size is up to 256 × 256, and propose to optimize the block-wise SAO filter by considering computation complexity and compression efficiency. Furthermore, we segment the band interval by equal probability of the sample intensity distributions, which facilitates the computation of better band offsets to attenuate ringing artifact due to quantization errors or encoded motion vectors. Experimental results show that the proposed superSAO method outperforms state-of-the-art approaches by obtaining 4.6% bandwidth reduction on average for the low delay and high-compression video applications.

Crowd counting is challenging due to unconstrained imaging factors, e.g., background clutters, non-uniform distribution of people, large scale and perspective variations. Dealing with these problems using deep neural networks requires rich prior knowledge and multi-scale contextual representations. In this paper, we propose a Cross-stage Refinement Network (CRNet) that can refine predicted density maps progressively based on hierarchical multi-level density priors. In particular, CRNet is composed of several fully convolutional networks. They are stacked together recursively with the previous output as the next input, and each of them serves to utilize previous density output to gradually correct prediction errors of crowd areas and refine the predicted density maps at different stages. Cross-stage multi-level density priors are further exploited in our recurrent framework by the cross-stage skip layers based on ConvLSTM. To cope with different challenges of unconstrained crowd scenes, we explore different crowd-specific data augmentation methods to mimic real-world scenarios and enrich crowd feature representations from different aspects. Extensive experiments show the proposed method achieves superior performances against state-of-the-art methods on four widely-used challenging benchmarks in terms of counting accuracy and density map quality. Code and models are available at this https://github.com/lytgftyf/Crowd-Counting-via-Cross-stage-Refinement-Networks.

Intracranial blood vessels segmentation from computed tomography angiography (CTA) volumes is a promising biomarker for diagnosis and therapeutic treatment in cerebrovascular diseases. These segmentation outputs are a fundamental requirement in the development of automated decision support systems for preoperative assessment or intraoperative guidance in neuropathology. The state-of-the-art in medical image segmentation methods are reliant on deep learning architectures based on convolutional neural networks. However, despite their popularity, there is a research gap in the current deep learning architectures optimized to address the technical challenges in blood vessel segmentation. These challenges include: (i) the extraction of concrete brain vessels close to the skull; and (ii) the precise marking of the vessel locations. We propose an Optimally Fused Fully end-to-end Network (OFF-eNET) for automatic segmentation of the volumetric 3D intracranial vascular structures. OFF-eNET comprises of three modules. In the first module, we exploit the up-skip connections to enhance information flow, and dilated convolution for detailed preservation of spatial feature map that are designed for thin blood vessels. In the second module, we employ residual mapping along with inception module for speedy network convergence and richer visual representation. For the third module, we make use of the transferred knowledge in the form of cascaded training strategy to gradually optimize the three segmentation stages (basic, complete, and enhanced) to segment thin vessels located close to the skull. All these modules are designed to be computationally efficient. Our OFF-eNET, evaluated using 70 CTA image volumes, resulted in 90.75% performance in the segmentation of intracranial blood vessels and outperformed the state-of-the-art counterparts.

Supervoxel segmentation algorithm has been applied as a preprocessing step for many vision tasks. However, existing supervoxel segmentation algorithms cannot generate hierarchical supervoxel segmentation well preserving the spatiotemporal boundaries in real time, which prevents the downstream applications from accurate and efficient processing. In this paper, we propose a real-time hierarchical supervoxel segmentation algorithm based on the minimum spanning tree (MST), which achieves state-of-the-art accuracy meanwhile at least 11× faster than existing methods. In particular, we present a dynamic graph updating operation into the iterative construction process of the MST, which can geometrically decrease the numbers of vertices and edges. In this way, the proposed method is able to generate arbitrary scales of supervoxels on the fly. We prove the efficiency of our algorithm that can produce hierarchical supervoxels in the time complexity of O(n) , where n denotes the number of voxels in the input video. Quantitative and qualitative evaluations on public benchmarks demonstrate that our proposed algorithm significantly outperforms the state-of-the-art algorithms in terms of supervoxel segmentation accuracy and computational efficiency. Furthermore, we demonstrate the effectiveness of the proposed method on a downstream application of video object segmentation.

Unsupervised domain adaptation has increasingly gained interest in medical image computing, aiming to tackle the performance degradation of deep neural networks when being deployed to unseen data with heterogeneous characteristics. In this work, we present a novel unsupervised domain adaptation framework, named as Synergistic Image and Feature Alignment (SIFA), to effectively adapt a segmentation network to an unlabeled target domain. Our proposed SIFA conducts synergistic alignment of domains from both image and feature perspectives. In particular, we simultaneously transform the appearance of images across domains and enhance domain-invariance of the extracted features by leveraging adversarial learning in multiple aspects and with a deeply supervised mechanism. The feature encoder is shared between both adaptive perspectives to leverage their mutual benefits via end-to-end learning. We have extensively evaluated our method with cardiac substructure segmentation and abdominal multi-organ segmentation for bidirectional cross-modality adaptation between MRI and CT images. Experimental results on two different tasks demonstrate that our SIFA method is effective in improving segmentation performance on unlabeled target images, and outperforms the state-of-the-art domain adaptation approaches by a large margin.

Lung cancer is the leading cause of cancer deaths worldwide and early diagnosis of lung nodule is of great importance for therapeutic treatment and saving lives. Automated lung nodule analysis requires both accurate lung nodule benign-malignant classification and attribute score regression. However, this is quite challenging due to the considerable difficulty of lung nodule heterogeneity modeling and the limited discrimination capability on ambiguous cases. To solve these challenges, we propose a Multi-Task deep model with Margin Ranking loss (referred as MTMR-Net) for automated lung nodule analysis. Compared to existing methods which consider these two tasks separately, the relatedness between lung nodule classification and attribute score regression is explicitly explored in a cause-and-effect manner within our multi-task deep model, which can contribute to the performance gains of both tasks. The results of different tasks can be yielded simultaneously for assisting the radiologists in diagnosis interpretation. Furthermore, a Siamese network with a margin ranking loss is elaborately designed to enhance the discrimination capability on ambiguous nodule cases. To further explore the internal relationship between two tasks and validate the effectiveness of the proposed model, we use the recursive feature elimination method to iteratively rank the most malignancy-related features. We validate the efficacy of our method MTMR-Net on the public benchmark LIDC-IDRI dataset. Extensive experiments show that the diagnosis results with internal relationship explicitly explored in our model has met some similar patterns in clinical usage and also demonstrate that our approach can achieve competitive classification performance and more accurate scoring on attributes over the state-of-the-arts. Codes are publicly available at: https://github.com/CaptainWilliam/MTMR-NET.

Sub-cortical brain structure segmentation is of great importance for diagnosing neuropsychiatric disorders. However, developing an automatic approach to segmenting sub-cortical brain structures remains very challenging due to the ambiguous boundaries, complex anatomical structures, and large variance of shapes. This paper presents a novel deep network architecture, namely Ψ -Net, for sub-cortical brain structure segmentation, aiming at selectively aggregating features and boosting the information propagation in a deep convolutional neural network (CNN). To achieve this, we first formulate a densely convolutional LSTM module (DC-LSTM) to selectively aggregate the convolutional features with the same spatial resolution at the same stage of a CNN. This helps to promote the discriminativeness of features at each CNN stage. Second, we stack multiple DC-LSTMs from the deepest stage to the shallowest stage to progressively enrich low-level feature maps with high-level context. We employ two benchmark datasets on sub-cortical brain structure segmentation, and perform various experiments to evaluate the proposed Ψ -Net. The experimental results show that our network performs favorably against the state-of-the-art methods on both benchmark datasets.

In High Efficiency Video Coding (HEVC), multiple-QP (quantization parameter) optimization can adapt to a local video content. However, the multiple-QP implementation in the HEVC reference software (HM 16.6) achieves the best QP value for each coding block with a large amount of computational complexity. To address this challenge, we propose a fast rate-constrained multiple-QP optimization approach for the HM platform. We first introduce a template-based transform coefficient selection method which can save the overall complexity of entropy coding. In addition, we model the multiple-QP determination as a new rate-constrained optimization problem, and finally, we get a feasible solution with a lower computation overhead. Experimental results show that our method dramatically reduces the average complexity under the all-intra, low-delay and random-access configuration.

Taking photos through a glass window leads to glare or reflection, which might distract the viewer from the scene behind the window. In this paper, we involve user interaction to tackle the ill-posedness of the reflection removal problem. Users are allowed to draw strokes or lassos to indicate the background and reflection layers. Instead of designing hand-crafted features, we propose the edge-aware cascaded networks for reflection removal. The proposed network is a two-stage pipeline. The first stage takes the edge hints converted from user guidance and the image with reflection as input, and then separates the input image into the background and reflection layers. The second stage involves a refinement network to recover the missing details of the background layers. We simulate different types of user guidance, and the networks are trained on simulated data. The cascaded networks are end-to-end and perform with a single feed-forward pass, enabling fast editing. Extensive experimental evaluations demonstrate that the proposed used-guided reflection removal network yields better performance than the state-of-the-art methods on real-world scenarios. Furthermore, we show that novice users can easily generate reflection-free images, and large improvements in reflection removal quality can be obtained in just one minute.

Top-down, goal-driven visual saliency exerts a huge influence on the human visual system for performing visual tasks. Text generations, like visual question answering (VQA) and visual question generation (VQG), have intrinsic connections with top-down saliency, which is usually involved in both VQA and VQG processes in an unsupervised manner. However, it is shown that the regions that humans choose to look at to answer questions are very different from the unsupervised attention models. In this brief, we aim to explore the intrinsic relationship between top-down saliency and text generations, and to figure out whether an accurate saliency response benefits text generation. To this end, we propose a dual supervised network with dynamic parameter prediction. Dual-supervision explicitly exploits the probabilistic correlation between the primal task top-down saliency detection and the dual task text generation, while dynamic parameter prediction encodes the given text (i.e., question or answer) into the fully convolutional network. Extensive experiments show the proposed top-down saliency method achieves the best correlation with human attention among various baselines. In addition, the proposed model can be guided by either questions or answers, and output the counterpart. Furthermore, we show that combining human-like visual question-saliency improves the performance of both answer and question generations.

Cloud storage security has gained considerable research efforts with the wide adoption of cloud computing. As a security mechanism, researchers have been investigating cloud storage auditing schemes that enable a user to verify whether the cloud keeps the user's outsourced data undamaged. However, existing schemes have usability issues in compatibility with existing real world cloud storage applications, error-tolerance, and efficiency. To mitigate this usability gap, this article proposes a new general cloud storage auditing scheme that is more usable. The proposed scheme uses the idea of integrating linear error correcting codes and linear homomorphic authentication schemes together. This integration uses only one additional block to achieve error tolerance and authentication simultaneously. To demonstrate the power of the general construction, we also propose one detailed scheme based on the proposed general construction using the Reed Solomon code and the universal hash based MAC authentication scheme, both of which are implemented over the computation-efficient Galois field GF(28). We also show that the proposed scheme is secure under the standard definition. Moreover, we implemented and open-sourced the proposed scheme. Experimental results show that the proposed scheme is orders of magnitude more efficient than the state-of-the-art scheme.

Common spatial pattern (CSP) is a classic method commonly used in multichannel electroencephalogram (EEG) signal processing, which aims to extract effective features for binary classification by solving spatial filters that maximize the ratio of filtered dispersion between two classes. The aim of this paper is to improve the performance of the conventional CSP method, which will be badly influenced by noises. The recently proposed quantized minimum error entropy (QMEE) criterion is applied to structure a new objective function instead of the L2-norm in the conventional CSP. Quantization is utilized to reduce the computational complexity. The new objective function is optimized by a gradient-based iterative algorithm. The desirable performance of the QMEE-based CSP method, namely CSP-QMEE, is demonstrated with a toy example and two real EEG datasets, including Dataset IIb of the brain-computer interfaces (BCIs) Competition IV (three channels) and Dataset IIIa of the BCI Competition III (60 channels). The new method can achieve satisfactory performance compared to existing methods on all datasets. The promising results in this paper suggest that the CSP-QMEE may become a powerful tool for BCIs.

Texture smoothing aims to smooth out textures in images, while retaining the prominent structures. This paper presents a saliency-aware approach to the problem with two key contributions. First, we design a deep saliency network with guided non-local blocks (GNLBs) for learning long-range pixel dependencies by taking the predicted saliency map at former layer as the guidance image to help suppress the non-saliency regions in the shallow layer. The GNLB computes the saliency response at a position by a weighted sum of features at all positions, and enables us to produce results that outperform existing deep saliency models. Second, we formulate a joint optimization framework to take saliency information when iteratively separating textures from structures: on the texture layer, we smooth out structures with the help of the saliency information and migrate structures from the texture to structure layer, while on the structure layer, we adopt another deep model to detect edges and simultaneous sparse coding to push textures back to the texture layer. We tested our method on a rich variety of images and compared it with several state-of-the-art methods. Both visual and quantitative comparison results show that our method better preserves structures while removing the texture components.

This paper presents a novel unsupervised domain adaptation framework, called Synergistic Image and Feature Adaptation (SIFA), to effectively tackle the problem of domain shift. Domain adaptation has become an important and hot topic in recent studies on deep learning, aiming to recover performance degradation when applying the neural networks to new testing domains. Our proposed SIFA is an elegant learning diagram which presents synergistic fusion of adaptations from both image and feature perspectives. In particular, we simultaneously transform the appearance of images across domains and enhance domain-invariance of the extracted features towards the segmentation task. The feature encoder layers are shared by both perspectives to grasp their mutual benefits during the end-to-end learning procedure. Without using any annotation from the target domain, the learning of our unified model is guided by adversarial losses, with multiple discriminators employed from various aspects. We have extensively validated our method with a challenging application of crossmodality medical image segmentation of cardiac structures. Experimental results demonstrate that our SIFA model recovers the degraded performance from 17.2% to 73.0%, and outperforms the state-of-the-art methods by a significant margin.

Due to the lack of paired data, the training of image reflection removal relies heavily on synthesizing reflection images. However, existing methods model reflection as a linear combination model, which cannot fully simulate the real-world scenarios. In this paper, we inject non-linearity into reflection removal from two aspects. First, instead of synthesizing reflection with a fixed combination factor or kernel, we propose to synthesize reflection images by predicting a non-linear alpha blending mask. This enables a free combination of different blurry kernels, leading to a controllable and diverse reflection synthesis. Second, we design a cascaded network for reflection removal with three tasks: predicting the transmission layer, reflection layer, and the non-linear alpha blending mask. The former two tasks are the fundamental outputs, while the latter one being the side output of the network. This side output, on the other hand, making the training a closed loop, so that the separated transmission and reflection layers can be recombined together for training with a reconstruction loss. Extensive quantitative and qualitative experiments demonstrate the proposed synthesis and removal approaches outperforms state-of-the-art methods on two standard benchmarks, as well as in real-world scenarios.

In 3D surface reconstruction from normals, discontinuity preservation is an important but challenging task. However, existing studies fail to address the discontinuous normal maps by enforcing the surface integrability in the continuous domain. This paper introduces a robust approach to preserve the surface discontinuity in the discrete geometry way. Firstly, we design two representative normal incompatibility features and propose an efficient discontinuity detection scheme to determine the splitting pattern for a discrete mesh. Secondly, we model the discontinuity preservation problem as a light-weight energy optimization framework by jointly considering the discontinuity detection and the overall reconstruction error. Lastly, we further shrink the feasible solution space to reduce the complexity based on the prior knowledge. Experiments show that the proposed method achieves the best performance on an extensive 3D dataset compared with the state-of-the-arts in terms of mean angular error and computational complexity.

This paper presents a deep multi-model fusion network to attentively integrate multiple models to separate layers and boost the performance in single-image dehazing. To do so, we first formulate the attentional feature integration module to maximize the integration of the convolutional neural network (CNN) features at different CNN layers and generate the attentional multi-level integrated features (AMLIF). Then, from the AMLIF, we further predict a haze-free result for an atmospheric scattering model, as well as for four haze-layer separation models, and then fuse the results together to produce the final haze-free image. To evaluate the effectiveness of our method, we compare our network with several state-of-the-art methods on two widely-used dehazing benchmark datasets, as well as on two sets of real-world hazy images. Experimental results demonstrate clear quantitative and qualitative improvements of our method over the state-of-the-arts.

In this paper, we present a novel and effective technique for short-term traffic flow forecasting. Our main contribution is an extension of Kalman filter, such that it becomes to be able to identify the noise and then filter out it; we hence named the present technique as noise-identified Kalman filter. Our epistemological perspective is that the classic Kalman filter filters out not only the noise but also useful signals. We hence develop the Kalman filter for de-noising while preserving the useful signals by devising a cost function. By conducting extensive experiments on four benchmark data sets, the proposed technique is firmly verified to be effective for short-term traffic flow forecasting, outperforming not only the classic Kalman filter but also other frequently-used parametric and non-parametric techniques.

Abstract Semantic surface decomposition (SSD) facilitates various geometry processing and product re‐design tasks. Filter‐based techniques are meaningful and widely used to achieve the SSD, which however often leads to surface either under‐fitting or over‐fitting. In this paper, we propose a reliable rolling‐guided point normal filtering method to decompose textures from a captured point cloud surface. Our method is built on the geometry assumption that 3D surfaces are comprised of an underlying shape (US) and a variety of bump ups and downs (BUDs) on the US. We have three core contributions. First, by considering the BUDs as surface textures, we present a RANSAC‐based sub‐neighborhood detection scheme to distinguish the US and the textures. Second, to better preserve the US (especially the prominent structures), we introduce a patch shift scheme to estimate the guidance normal for feeding the rolling‐guided filter. Third, we formulate a new position updating scheme to alleviate the common uneven distribution of points. Both visual and numerical experiments demonstrate that our method is comparable to state‐of‐the‐art methods in terms of the robustness of texture removal and the effectiveness of the underlying shape preservation.

Abstract This paper addresses the nontraditional but practically meaningful reversibility problem of mesh filtering. This reverse‐filtering approach (termed a DeFilter) seeks to recover the geometry of a set of filtered meshes to their artifact‐free status. To solve this scenario, we adapt cascaded normal regression (CNR) to understand the commonly used mesh filters and recover automatically the mesh geometry that was lost through various geometric operations. We formulate mesh defiltering by an extreme learning machine (ELM) on the mesh normals at an offline training stage and perform it automatically at a runtime defiltering stage. Specifically, (1) to measure the local geometry of a filtered mesh, we develop a generalized reverse Filtered Facet Normal Descriptor (grFND) in the consistent neighbors; (2) to map the grFNDs to the normals of the ground‐truth meshes, we learn a regression function from a set of filtered meshes and their ground‐truth counterparts; and (3) at runtime, we reversely filter the normals of a filtered mesh, using the learned regression function for recovering the lost geometry. We evaluate multiple quantitative and qualitative results on synthetic and real data to verify our DeFilter's performance thoroughly. From a practical point of view, our method can recover the lost geometry of denoised meshes without needing to know the exact filter used previously, and can act as a geometry‐recovery plugin for most of the state‐of‐the‐art methods of mesh denoising.

Linear regression model (LRM) based on mean square error (MSE) criterion is widely used in Granger causality analysis (GCA), which is the most commonly used method to detect the causality between a pair of time series. However, when signals are seriously contaminated by non-Gaussian noises, the LRM coefficients will be inaccurately identified. This may cause the GCA to detect a wrong causal relationship. Minimum error entropy (MEE) criterion can be used to replace the MSE criterion to deal with the non-Gaussian noises. But its calculation requires a double summation operation, which brings computational bottlenecks to GCA especially when sizes of the signals are large. To address the aforementioned problems, in this letter, we propose a new method called GCA based on the quantized MEE (QMEE) criterion (GCA-QMEE), in which the QMEE criterion is applied to identify the LRM coefficients and the quantized error entropy is used to calculate the causality indexes. Compared with the traditional GCA, the proposed GCA-QMEE not only makes the results more discriminative, but also more robust. Its computational complexity is also not high because of the quantization operation. Illustrative examples on synthetic and EEG datasets are provided to verify the desirable performance and the availability of the GCA-QMEE.

In order to efficiently utilize the information in the data and eliminate the negative effects of outliers in the principal component analysis (PCA) method, in this paper, we propose a novel robust sparse PCA method based on maximum correntropy criterion (MCC) with high-order manifold constraints called the RHSPCA. Compared with the traditional PCA methods, the proposed RHSPCA has the following benefits: 1) the MCC regression term is more robust to outliers than the MSE-based regression term; 2) thanks to the high-order manifold constraints, the low-dimensional representations can preserve the local relations of the data and greatly improve the clustering and classification performance for image processing tasks; and 3) in order to further counteract the adverse effects of outliers, the MCC-based samples' mean is proposed to better centralize the data. We also propose a new solver based on the half-quadratic technique and accelerated block coordinate update strategy to solve the RHSPCA model. Extensive experimental results show that the proposed method can outperform the state-of-the-art robust PCA methods on a variety of image processing tasks, including reconstruction, clustering, and classification, on outliers contaminated datasets.

The retinal vessel is one of the determining factors in an ophthalmic examination. Automatic extraction of retinal vessels from low-quality retinal images still remains a challenging problem. In this paper, we propose a robust and effective approach that qualitatively improves the detection of low-contrast and narrow vessels. Rather than using the pixel grid, we use a superpixel as the elementary unit of our vessel segmentation scheme. We regularize this scheme by combining the geometrical structure, texture, color, and space information in the superpixel graph. And the segmentation results are then refined by employing the efficient minimum spanning superpixel tree to detect and capture both global and local structure of the retinal images. Such an effective and structure-aware tree detector significantly improves the detection around the pathologic area. Experimental results have shown that the proposed technique achieves advantageous connectivity-area-length (CAL) scores of 80.92% and 69.06% on two public datasets, namely, DRIVE and STARE, thereby outperforming state-of-the-art segmentation methods. In addition, the tests on the challenging retinal image database have further demonstrated the effectiveness of our method. Our approach achieves satisfactory segmentation performance in comparison with state-of-the-art methods. Our technique provides an automated method for effectively extracting the vessel from fundus images.

Parkinson's disease (PD) is a neurodegenerative progressive disease that mainly affects the motor systems of patients. To slow this disease deterioration, early and accurate diagnosis of PD is an effective way, which alleviates mental and physical sufferings by clinical intervention. In this paper, we propose a joint regression and classification framework for PD diagnosis via magnetic resonance and diffusion tensor imaging data. Specifically, we devise a unified multitask feature selection model to explore multiple relationships among features, samples, and clinical scores. We regress four clinical variables of depression, sleep, olfaction, cognition scores, as well as perform the classification of PD disease from the multimodal data. The multitask model explores the relationships at the level of clinical scores, image features, and subjects, to select the most informative and diseased-related features for diagnosis. The proposed method is evaluated on the public Parkinson's progression markers initiative dataset. The extensive experimental results show that the multitask framework can effectively boost the performance of regression and classification and outperforms other state-of-the-art methods. The computerized predictions of clinical scores and label for PD diagnosis may offer quantitative reference for decision support as well.

Vision-based vehicle detection approaches achieve incredible success in recent years with the development of deep convolutional neural network (CNN). However, existing CNN-based algorithms suffer from the problem that the convolutional features are scale-sensitive in object detection task but it is common that traffic images and videos contain vehicles with a large variance of scales. In this paper, we delve into the source of scale sensitivity, and reveal two key issues: 1) existing RoI pooling destroys the structure of small scale objects and 2) the large intra-class distance for a large variance of scales exceeds the representation capability of a single network. Based on these findings, we present a scale-insensitive convolutional neural network (SINet) for fast detecting vehicles with a large variance of scales. First, we present a context-aware RoI pooling to maintain the contextual information and original structure of small scale objects. Second, we present a multi-branch decision network to minimize the intra-class distance of features. These lightweight techniques bring zero extra time complexity but prominent detection accuracy improvement. The proposed techniques can be equipped with any deep network architectures and keep them trained end-to-end. Our SINet achieves state-of-the-art performance in terms of accuracy and speed (up to 37 FPS) on the KITTI benchmark and a new highway dataset, which contains a large variance of scales and extremely small objects.

We present a novel approach for segmenting overlapping cytoplasm of cells in cervical smear images by leveraging the adaptive shape priors extracted from cytoplasm's contour fragments and shape statistics. The main challenge of this task is that many occluded boundaries in cytoplasm clumps are extremely difficult to be identified and, sometimes, even visually indistinguishable. Given a clump where multiple cytoplasms overlap, our method starts by cutting its contour into a set of contour fragments. We then locate the corresponding contour fragments of each cytoplasm by a grouping process. For each cytoplasm, according to the grouped fragments and a set of known shape references, we construct its shape and, then, connect the fragments to form a closed contour as the segmentation result, which is explicitly constrained by the constructed shape. We further integrate the intensity and curvature information, which is complementary to the shape priors extracted from contour fragments, into our framework to improve the segmentation accuracy. We propose to iteratively conduct fragments grouping, shape constructing, and fragments connecting for progressively refining the shape priors and improving the segmentation results. We extensively evaluate the effectiveness of our method on two typical cervical smear datasets. The experimental results demonstrate that our approach is highly effective and consistently outperforms the state-of-the-art approaches. The proposed method is general enough to be applied to other similar microscopic image segmentation tasks, where heavily overlapped objects exist.

Automatic prostate segmentation in transrectal ultrasound (TRUS) images is of essential importance for image-guided prostate interventions and treatment planning. However, developing such automatic solutions remains very challenging due to the missing/ambiguous boundary and inhomogeneous intensity distribution of the prostate in TRUS, as well as the large variability in prostate shapes. This paper develops a novel 3D deep neural network equipped with attention modules for better prostate segmentation in TRUS by fully exploiting the complementary information encoded in different layers of the convolutional neural network (CNN). Our attention module utilizes the attention mechanism to selectively leverage the multi-level features integrated from different layers to refine the features at each individual layer, suppressing the non-prostate noise at shallow layers of the CNN and increasing more prostate details into features at deep layers. Experimental results on challenging 3D TRUS volumes show that our method attains satisfactory segmentation performance. The proposed attention mechanism is a general strategy to aggregate multi-level deep features and has the potential to be used for other medical image segmentation tasks. The code is publicly available at https://github.com/wulalago/DAF3D.

Volumetric ultrasound is rapidly emerging as a viable imaging modality for routine prenatal examinations. Biometrics obtained from the volumetric segmentation shed light on the reformation of precise maternal and fetal health monitoring. However, the poor image quality, low contrast, boundary ambiguity, and complex anatomy shapes conspire toward a great lack of efficient tools for the segmentation. It makes 3-D ultrasound difficult to interpret and hinders the widespread of 3-D ultrasound in obstetrics. In this paper, we are looking at the problem of semantic segmentation in prenatal ultrasound volumes. Our contribution is threefold: 1) we propose the first and fully automatic framework to simultaneously segment multiple anatomical structures with intensive clinical interest, including fetus, gestational sac, and placenta, which remains a rarely studied and arduous challenge; 2) we propose a composite architecture for dense labeling, in which a customized 3-D fully convolutional network explores spatial intensity concurrency for initial labeling, while a multi-directional recurrent neural network (RNN) encodes spatial sequentiality to combat boundary ambiguity for significant refinement; and 3) we introduce a hierarchical deep supervision mechanism to boost the information flow within RNN and fit the latent sequence hierarchy in fine scales, and further improve the segmentation results. Extensively verified on in-house large data sets, our method illustrates a superior segmentation performance, decent agreements with expert measurements and high reproducibilities against scanning variations, and thus is promising in advancing the prenatal ultrasound examinations.

Mesh denoising is a classical, yet not well-solved problem in digital geometry processing. The challenge arises from noise removal with the minimal disturbance of surface intrinsic properties (e.g., sharp features and shallow details). We propose a new patch normal co-filter (PcFilter) for mesh denoising. It is inspired by the geometry statistics which show that surface patches with similar intrinsic properties exist on the underlying surface of a noisy mesh. We model the PcFilter as a low-rank matrix recovery problem of similar-patch collaboration, aiming at removing different levels of noise, yet preserving various surface features. We generalize our model to pursue the low-rank matrix recovery in the kernel space for handling the nonlinear structure contained in the data. By making use of the block coordinate descent minimization and the specifics of a proximal based coordinate descent method, we optimize the nonlinear and nonconvex objective function efficiently. The detailed quantitative and qualitative results on synthetic and real data show that the PcFilter competes favorably with the state-of-the-art methods in surface accuracy and noise-robustness.

Bas-relief is characterized by its unique presentation of intrinsic shape properties and/or detailed appearance using materials raised up in different degrees above a background. However, many bas-relief modeling methods could not manipulate scene details well. We propose a simple and effective solution for two kinds of bas-relief modeling (i.e., structure-preserving and detail-preserving) which is different from the prior tone mapping alike methods. Our idea originates from an observation on typical 3D models, which are decomposed into a piecewise smooth base layer and a detail layer in normal field. Proper manipulation of the two layers contributes to both structure-preserving and detail-preserving bas-relief modeling. We solve the modeling problem in a discrete geometry processing setup that uses normal-based mesh processing as a theoretical foundation. Specifically, using the two-step mesh smoothing mechanism as a bridge, we transfer the bas-relief modeling problem into a discrete space, and solve it in a least-squares manner. Experiments and comparisons to other methods show that (i) geometry details are better preserved in the scenario with high compression ratios, and (ii) structures are clearly preserved without shape distortion and interference from details.

Salient object detection is a fundamental yet challenging problem in computer vision, aiming to highlight the most visually distinctive objects or regions in an image. Recent works benefit from the development of fully convolutional neural networks (FCNs) and achieve great success by integrating features from multiple layers of FCNs. However, the integrated features tend to include non-salient regions (due to low level features of the FCN) or lost details of salient objects (due to high level features of the FCN) when producing the saliency maps. In this paper, we develop a novel deep saliency network equipped with recurrently aggregated deep features (RADF) to more accurately detect salient objects from an image by fully exploiting the complementary saliency information captured in different layers. The RADF utilizes the multi-level features integrated from different layers of a FCN to recurrently refine the features at each layer, suppressing the non-salient noise at low-level of the FCN and increasing more salient details into features at high layers. We perform experiments to evaluate the effectiveness of the proposed network on 5 famous saliency detection benchmarks and compare it with 15 state-of-the-art methods. Our method ranks first in 4 of the 5 datasets and second in the left dataset.

We present a novel method for automated segmentation of overlapping cytoplasm in cervical smear images based on contour fragments. We formulate the segmentation problem as a graphical model, and employ the contour fragments generated from cytoplasm clump to construct the graph. Compared with traditional methods that are based on pixels, our contour fragment-based solution can take more geometric information into account and hence generate more accurate prediction of the overlapping boundaries. We further design a novel energy function for the graph, and by minimizing the energy function, fragments that come from the same cytoplasm are selected into the same set. To construct the energy function, our fragments-based data term and pairwise term are measured from the spatial relation and shape prior, which offer more geometric information for the occluded boundary inference. Afterwards, occluded boundaries are inferred using the minimal path model, in which shape of each individual cytoplasm is reconstructed on the selected fragments set. Constructed shape is used as a constraint to locate the searching area, and curvature regulation is enforced to promote the smoothness of inference result. The inference result, in turn, is used as the shape prior to construct a high-level shape regulation energy term of the built graph, and then graph energy is updated. In other words, fragments selection and occluded boundary inference are iterative processed; this interaction makes more potential shape information accessible. Using two cervical smear datasets, the performance of our method is extensively evaluated and compared with that of the state-of-the-art approaches; the results show the superiority of the proposed method.

Cell nuclei detection and fine-grained classification have been fundamental yet challenging problems in histopathology image analysis. Due to the nuclei tiny size, significant inter-/intra-class variances, as well as the inferior image quality, previous automated methods would easily suffer from limited accuracy and robustness. In the meanwhile, existing approaches usually deal with these two tasks independently, which would neglect the close relatedness of them. In this paper, we present a novel method of sibling fully convolutional network with prior objectness interaction (called SFCN-OPI) to tackle the two tasks simultaneously and interactively using a unified end-to-end framework. Specifically, the sibling FCN branches share features in earlier layers while holding respective higher layers for specific tasks. More importantly, the detection branch outputs the objectness prior which dynamically interacts with the fine-grained classification sibling branch during the training and testing processes. With this mechanism, the fine-grained classification successfully focuses on regions with high confidence of nuclei existence and outputs the conditional probability, which in turn benefits the detection through back propagation. Extensive experiments on colon cancer histology images have validated the effectiveness of our proposed SFCN-OPI and our method has outperformed the state-of-the-art methods by a large margin.

Shadow detection is a fundamental and challenging task, since it requires an understanding of global image semantics and there are various backgrounds around shadows. This paper presents a novel network for shadow detection by analyzing image context in a direction-aware manner. To achieve this, we first formulate the direction-aware attention mechanism in a spatial recurrent neural network (RNN) by introducing attention weights when aggregating spatial context features in the RNN. By learning these weights through training, we can recover direction-aware spatial context (DSC) for detecting shadows. This design is developed into the DSC module and embedded in a CNN to learn DSC features at different levels. Moreover, a weighted cross entropy loss is designed to make the training more effective. We employ two common shadow detection benchmark datasets and perform various experiments to evaluate our network. Experimental results show that our network outperforms state-of-the-art methods and achieves 97% accuracy and 38% reduction on balance error rate.

Saliency detection is a fundamental yet challenging task in computer vision, aiming at highlighting the most visually distinctive objects in an image. We propose a novel recurrent residual refinement network (R^3Net) equipped with residual refinement blocks (RRBs) to more accurately detect salient regions of an input image. Our RRBs learn the residual between the intermediate saliency prediction and the ground truth by alternatively leveraging the low-level integrated features and the high-level integrated features of a fully convolutional network (FCN). While the low-level integrated features are capable of capturing more saliency details, the high-level integrated features can reduce non-salient regions in the intermediate prediction. Furthermore, the RRBs can obtain complementary saliency information of the intermediate prediction, and add the residual into the intermediate prediction to refine the saliency maps. We evaluate the proposed R^3Net on five widely-used saliency detection benchmarks by comparing it with 16 state-of-the-art saliency detectors. Experimental results show that our network outperforms our competitors in all the benchmark datasets.

Lymph node metastasis is one of the most significant diagnostic indicators in breast cancer, which is traditionally observed under the microscope by pathologists. In recent years, computerized histology diagnosis has become one of the most rapidly expanding directions in the field of medical image computing, which aims to alleviate pathologists' workload and simultaneously reduce misdiagnosis rate. However, automatic detection of lymph node metastases from whole slide images remains a challenging problem, due to the large-scale data with enormous resolutions and existence of hard mimics resulting in a large number of false positives. In this paper, we propose a novel framework by leveraging fully convolutional networks for efficient inference to meet the speed requirement for clinical practice, while reconstructing dense predictions under different offsets for ensuring accurate detection on both microand macro-metastases. Incorporating with the strategies of asynchronous sample prefetching and hard negative mining, the network can be effectively trained. Extensive experiments on the benchmark dataset of 2016 Camelyon Grand Challenge corroborated the efficacy of our method. Compared with the state-of-the-art methods, our method achieved superior performance with a faster speed on the tumor localization task and even surpassed human performance on the WSI classification task.

Correntropy is a second order statistical measure in kernel space, which has been successfully applied in robust learning and signal processing. In this paper, we define a nonsecond order statistical measure in kernel space, called the kernel mean- power error (KMPE), including the correntropic loss (C-Loss) as a special case. Some basic properties of KMPE are presented. In particular, we apply the KMPE to extreme learning machine (ELM) and principal component analysis (PCA), and develop two robust learning algorithms, namely ELM-KMPE and PCA-KMPE. Experimental results on synthetic and benchmark data show that the developed algorithms can achieve better performance when compared with some existing methods.

Verifiable pattern matching is the problem of finding a given pattern verifiably from the outsourced textual data, which is resident in an untrusted remote server. This problem has drawn much attention due to a large number of applications. The state-of-the-art method for this problem suffers from low efficiency. To enable fast verifiable pattern matching, we propose a novel scheme in this paper. Our scheme is based on an ordered set accumulator data structure and a newly developed verifiable suffix array structure, which only involves fast cryptographic hash computations. Our scheme also supports fast multiple-occurrence pattern matching. A striking feature of our proposed scheme is that our scheme works even with no secret keys, which ensures public verifiability. We conduct extensive experiments to evaluate the proposed scheme using Java. The results show that our scheme is orders of magnitude faster than the state-of-the-art work. Specifically, our scheme with public verifiability only costs a preprocessing time of 47 s (merely one-time off-line cost during outsourcing), a search time of 30 μs, a verification time of 149 μs, and a proof size of 2760 bytes for a verifiable pattern matching query with pattern length 200 on 10-million long textual data which consists of sequences of two-byte, Unicode characters in Java.

Head circumference (HC) is one of the most important biometrics in assessing fetal growth during prenatal ultrasound examinations. However, the manual measurement of this biometric by doctors often requires substantial experience. We developed a learning-based framework that used prior knowledge and employed a fast ellipse fitting method (ElliFit) to measure HC automatically. We first integrated the prior knowledge about the gestational age and ultrasound scanning depth into a random forest classifier to localize the fetal head. We further used phase symmetry to detect the center line of the fetal skull and employed ElliFit to fit the HC ellipse for measurement. The experimental results from 145 HC images showed that our method had an average measurement error of 1.7 mm and outperformed traditional methods. The experimental results demonstrated that our method shows great promise for applications in clinical practice.

Ultrasound imaging has become a prevalent examination method in prenatal diagnosis. Accurate acquisition of fetal facial standard plane (FFSP) is the most important precondition for subsequent diagnosis and measurement. In the past few years, considerable effort has been devoted to FFSP recognition using various hand-crafted features, but the recognition performance is still unsatisfactory due to the high intraclass variation of FFSPs and the high degree of visual similarity between FFSPs and other non-FFSPs. To improve the recognition performance, we propose a method to automatically recognize FFSP via a deep convolutional neural network (DCNN) architecture. The proposed DCNN consists of 16 convolutional layers with small 3 × 3 size kernels and three fully connected layers. A global average pooling is adopted in the last pooling layer to significantly reduce network parameters, which alleviates the overfitting problems and improves the performance under limited training data. Both the transfer learning strategy and a data augmentation technique tailored for FFSP are implemented to further boost the recognition performance. Extensive experiments demonstrate the advantage of our proposed method over traditional approaches and the effectiveness of DCNN to recognize FFSP for clinical diagnosis.

We propose an analysis of surgical videos that is based on a novel recurrent convolutional network (SV-RCNet), specifically for automatic workflow recognition from surgical videos online, which is a key component for developing the context-aware computer-assisted intervention systems. Different from previous methods which harness visual and temporal information separately, the proposed SV-RCNet seamlessly integrates a convolutional neural network (CNN) and a recurrent neural network (RNN) to form a novel recurrent convolutional architecture in order to take full advantages of the complementary information of visual and temporal features learned from surgical videos. We effectively train the SV-RCNet in an end-to-end manner so that the visual representations and sequential dynamics can be jointly optimized in the learning process. In order to produce more discriminative spatio-temporal features, we exploit a deep residual network (ResNet) and a long short term memory (LSTM) network, to extract visual features and temporal dependencies, respectively, and integrate them into the SV-RCNet. Moreover, based on the phase transition-sensitive predictions from the SV-RCNet, we propose a simple yet effective inference scheme, namely the prior knowledge inference (PKI), by leveraging the natural characteristic of surgical video. Such a strategy further improves the consistency of results and largely boosts the recognition performance. Extensive experiments have been conducted with the MICCAI 2016 Modeling and Monitoring of Computer Assisted Interventions Workflow Challenge dataset and Cholec80 dataset to validate SV-RCNet. Our approach not only achieves superior performance on these two datasets but also outperforms the state-of-the-art methods by a significant margin.

Realistic liquid animations with thin sheets or streams are crucial for creating fluid effects in digital media. However, it is challenging to simulate these appealing thin sheets or streams in the framework of smoothed particle hydrodynamics (SPH). The underlying reason for this challenge mainly lies in the inherent numerical instability of SPH due to inconsistent kernel interpolation, which is caused by the incomplete kernel support on the free surface and the particles' disorder dispersion within the simulation domain. To address this challenge, we propose a novel and effective approach to ensure the consistency of kernel interpolation at both internal flow and the free surface during the simulation such that these thin features can always be well maintained. First, we introduce a transport-velocity formulation to alleviate the disorder dispersion in the liquid domain. However, this formulation can only work in the internal flow, and it fails at the free surface because it cannot accurately estimate the density of particles there. To this end, we propose adaptively correcting the underestimated density caused by the incomplete kernel support of free-surface particles, which are identified by a geometry-aware anisotropic kernel, to counteract the inconsistent interpolation on the free surface. Then, we propose a novel scheme to further filter the background pressure to enhance the interactions between the internal flow and the free surface, as well as liquid and solid, such that the thin features generated from such interactions can be realistically simulated. The proposed approach can also achieve anticlumping and regularization effects in the entire simulation domain and, hence, further enhance the thin features in liquids. We evaluate our method on a variety of benchmark examples, and the results demonstrate that our method can achieve more appealing visual effects than state-of-the-art methods by realistically simulating more vivid thin features.

Turbulent vortices in smoke flows are crucial for a visually interesting appearance. Unfortunately, it is challenging to efficiently simulate these appealing effects in the framework of vortex filament methods. The vortex filaments in grids scheme allows to efficiently generate turbulent smoke with macroscopic vortical structures, but suffers from the projection-related dissipation, and thus the small-scale vortical structures under grid resolution are hard to capture. In addition, this scheme cannot be applied in wall-bounded turbulent smoke simulation, which requires efficiently handling smoke-obstacle interaction and creating vorticity at the obstacle boundary. To tackle above issues, we propose an effective filament-mesh particle-particle (FMPP) method for fast wall-bounded turbulent smoke simulation with ample details. The Filament-Mesh component approximates the smooth long-range interactions by splatting vortex filaments on grid, solving the Poisson problem with a fast solver, and then interpolating back to smoke particles. The Particle-Particle component introduces smoothed particle hydrodynamics (SPH) turbulence model for particles in the same grid, where interactions between particles cannot be properly captured under grid resolution. Then, we sample the surface of obstacles with boundary particles, allowing the interaction between smoke and obstacle being treated as pressure forces in SPH. Besides, the vortex formation region is defined at the back of obstacles, providing smoke particles flowing by the separation particles with a vorticity force to simulate the subsequent vortex shedding phenomenon. The proposed approach can synthesize the lost small-scale vortical structures and also achieve the smoke-obstacle interaction with vortex shedding at obstacle boundaries in a lightweight manner. The experimental results demonstrate that our FMPP method can achieve more appealing visual effects than vortex filaments in grids scheme by efficiently simulating more vivid thin turbulent features.

Boundary incompleteness raises great challenges to automatic prostate segmentation in ultrasound images. Shape prior can provide strong guidance in estimating the missing boundary, but traditional shape models often suffer from hand-crafted descriptors and local information loss in the fitting procedure. In this paper, we attempt to address those issues with a novel framework. The proposed framework can seamlessly integrate feature extraction and shape prior exploring, and estimate the complete boundary with a sequential manner. Our framework is composed of three key modules. Firstly, we serialize the static 2D prostate ultrasound images into dynamic sequences and then predict prostate shapes by sequentially exploring shape priors. Intuitively, we propose to learn the shape prior with the biologically plausible Recurrent Neural Networks (RNNs). This module is corroborated to be effective in dealing with the boundary incompleteness. Secondly, to alleviate the bias caused by different serialization manners, we propose a multi-view fusion strategy to merge shape predictions obtained from different perspectives. Thirdly, we further implant the RNN core into a multiscale Auto-Context scheme to successively refine the details of the shape prediction map. With extensive validation on challenging prostate ultrasound images, our framework bridges severe boundary incompleteness and achieves the best performance in prostate boundary delineation when compared with several advanced methods. Additionally, our approach is general and can be extended to other medical image segmentation tasks, where boundary incompleteness is one of the main challenges.

Automated prostate segmentation from 3D MR images is very challenging due to large variations of prostate shape and indistinct prostate boundaries. We propose a novel volumetric convolutional neural network (ConvNet) with mixed residual connections to cope with this challenging problem. Compared with previous methods, our volumetric ConvNet has two compelling advantages. First, it is implemented in a 3D manner and can fully exploit the 3D spatial contextual information of input data to perform efficient, precise and volume-to-volume prediction. Second and more important, the novel combination of residual connections (i.e., long and short) can greatly improve the training efficiency and discriminative capability of our network by enhancing the information propagation within the ConvNet both locally and globally. While the forward propagation of location information can improve the segmentation accuracy, the smooth backward propagation of gradient flow can accelerate the convergence speed and enhance the discrimination capability. Extensive experiments on the open MICCAI PROMISE12 challenge dataset corroborated the effectiveness of the proposed volumetric ConvNet with mixed residual connections. Our method ranked the first in the challenge, outperforming other competitors by a large margin with respect to most of evaluation metrics. The proposed volumetric ConvNet is general enough and can be easily extended to other medical image analysis tasks, especially ones with limited training data.

Motor imagery (MI) based on brain computer interfaces (BCIs) have been widely applied for upper limb motor rehabilitation. Due to the fact that a large number of disabled people need to restore or improve walking ability, it is also important to investigate the use of MI-based BCIs for lower limb motor rehabilitation. The brain activity of lower limb MI is more difficult to detect because of low reliability. The purpose of this study is to find a suitable paradigm of walking imagery to achieve better training effect and ensure reliable brain activity. We developed the text-based paradigm and the virtual environment (VE)-based paradigm, and evaluated their performance on identifying walking imagery from idle state.The experimental results provide evidences that the VE-based paradigm could improve the average classification accuracy. This paradigm would induce EEG patterns that make them easier for single-trial detection of walking imagery. This study has the potential to improve the reliability and robustness of walking imagery based BCIs.

Three-dimensional (3D) ultrasound is extensively used in obstetrics and gynecology, and realistic rendering results can both enhance the accuracy of diagnosis and facilitate communication between obstetricians and pregnant women. This paper proposes an interactive and realistic rendering method using global illumination effects for 3D ultrasound images with low signal-to-noise-ratio (SNR) values. The method is capable of generating both single and multiple scattering effects, and separately rendering these two illuminations using ray-casting. Volumetric photon mapping is used to simulate indirect illumination. In order to avoid complex storage structures and accelerate the computational speed, a new screen-space destiny estimation is proposed to calculate the radiance of each photon. In addition, to solve the issue of the foetal skin exceeding the dynamic range capability of the display device, high-dynamic range (HDR) methods are further incorporated in our method. Experiments demonstrate that our method can produce realistic rendering results with enhanced depth information compared to traditional approaches.

Virtual surgery is a serious game which provides an opportunity to acquire cognitive and technical surgical skills via virtual surgical training and planning. However, interactively and realistically manipulating the human organ and simulating its motion under interaction is still a challenging task in this field. The underlying reason for this issue is the conflict requirements for physical constraints with high fidelity and real-time performance. To achieve realistic simulation of human organ motion with volume conservation, smooth interpolation under large deformation and precise frictional contact mechanics of global behavior in surgical scenario. This paper presents a novel and effective patch Green coordinates based interpolation for embedded deformable model to achieve the volume-preserving and smooth interpolation effects. Besides, we resolve the frictional contact mechanics for embedded deformable model, and further provide the precise boundary conditions for mechanical solver. In addition, our embedded deformable model is based on the total lagrangian explicit dynamics (TLED) finite element method (FEM) solver, which can well handle the large biological tissue deformation with both nonlinear geometric and material properties. In real compression experiments, our method can achieve liver deformation with average accuracy of 3.02 mm. Besides, the experimental results demonstrate that our method can also achieve smoother interpolation and volume-preserving effects than original embedded deformable model, and allows complex and accurate organ motion with mechanical interactions in virtual surgery.

Ultrasound (US) imaging is a widely used screening tool for obstetric examination and diagnosis. Accurate acquisition of fetal standard planes with key anatomical structures is very crucial for substantial biometric measurement and diagnosis. However, the standard plane acquisition is a labor-intensive task and requires operator equipped with a thorough knowledge of fetal anatomy. Therefore, automatic approaches are highly demanded in clinical practice to alleviate the workload and boost the examination efficiency. The automatic detection of standard planes from US videos remains a challenging problem due to the high intraclass and low interclass variations of standard planes, and the relatively low image quality. Unlike previous studies which were specifically designed for individual anatomical standard planes, respectively, we present a general framework for the automatic identification of different standard planes from US videos. Distinct from conventional way that devises hand-crafted visual features for detection, our framework explores in- and between-plane feature learning with a novel composite framework of the convolutional and recurrent neural networks. To further address the issue of limited training data, a multitask learning framework is implemented to exploit common knowledge across detection tasks of distinctive standard planes for the augmentation of feature learning. Extensive experiments have been conducted on hundreds of US fetus videos to corroborate the better efficacy of the proposed framework on the difficult standard plane detection problem.

Automated polyp detection in colonoscopy videos has been demonstrated to be a promising way for colorectal cancer prevention and diagnosis. Traditional manual screening is time consuming, operator dependent, and error prone; hence, automated detection approach is highly demanded in clinical practice. However, automated polyp detection is very challenging due to high intraclass variations in polyp size, color, shape, and texture, and low interclass variations between polyps and hard mimics. In this paper, we propose a novel offline and online three-dimensional (3-D) deep learning integration framework by leveraging the 3-D fully convolutional network (3D-FCN) to tackle this challenging problem. Compared with the previous methods employing hand-crafted features or 2-D convolutional neural network, the 3D-FCN is capable of learning more representative spatio-temporal features from colonoscopy videos, and hence has more powerful discrimination capability. More importantly, we propose a novel online learning scheme to deal with the problem of limited training data by harnessing the specific information of an input video in the learning process. We integrate offline and online learning to effectively reduce the number of false positives generated by the offline network and further improve the detection performance. Extensive experiments on the dataset of MICCAI 2015 Challenge on Polyp Detection demonstrated the better performance of our method when compared with other competitors.

The gap between the computational and semantic features is the one of major factors that bottlenecks the computer-aided diagnosis (CAD) performance from clinical usage. To bridge this gap, we exploit three multi-task learning (MTL) schemes to leverage heterogeneous computational features derived from deep learning models of stacked denoising autoencoder (SDAE) and convolutional neural network (CNN), as well as hand-crafted Haar-like and HoG features, for the description of 9 semantic features for lung nodules in CT images. We regard that there may exist relations among the semantic features of "spiculation", "texture", "margin", etc., that can be explored with the MTL. The Lung Image Database Consortium (LIDC) data is adopted in this study for the rich annotation resources. The LIDC nodules were quantitatively scored w.r.t. 9 semantic features from 12 radiologists of several institutes in U.S.A. By treating each semantic feature as an individual task, the MTL schemes select and map the heterogeneous computational features toward the radiologists' ratings with cross validation evaluation schemes on the randomly selected 2400 nodules from the LIDC dataset. The experimental results suggest that the predicted semantic scores from the three MTL schemes are closer to the radiologists' ratings than the scores from single-task LASSO and elastic net regression methods. The proposed semantic attribute scoring scheme may provide richer quantitative assessments of nodules for better support of diagnostic decision and management. Meanwhile, the capability of the automatic association of medical image contents with the clinical semantic terms by our method may also assist the development of medical search engine.

Automated melanoma recognition in dermoscopy images is a very challenging task due to the low contrast of skin lesions, the huge intraclass variation of melanomas, the high degree of visual similarity between melanoma and non-melanoma lesions, and the existence of many artifacts in the image. In order to meet these challenges, we propose a novel method for melanoma recognition by leveraging very deep convolutional neural networks (CNNs). Compared with existing methods employing either low-level hand-crafted features or CNNs with shallower architectures, our substantially deeper networks (more than 50 layers) can acquire richer and more discriminative features for more accurate recognition. To take full advantage of very deep networks, we propose a set of schemes to ensure effective training and learning under limited training data. First, we apply the residual learning to cope with the degradation and overfitting problems when a network goes deeper. This technique can ensure that our networks benefit from the performance gains achieved by increasing network depth. Then, we construct a fully convolutional residual network (FCRN) for accurate skin lesion segmentation, and further enhance its capability by incorporating a multi-scale contextual information integration scheme. Finally, we seamlessly integrate the proposed FCRN (for segmentation) and other very deep residual networks (for classification) to form a two-stage framework. This framework enables the classification network to extract more representative and specific features based on segmented results instead of the whole dermoscopy images, further alleviating the insufficiency of training data. The proposed framework is extensively evaluated on ISBI 2016 Skin Lesion Analysis Towards Melanoma Detection Challenge dataset. Experimental results demonstrate the significant performance gains of the proposed framework, ranking the first in classification and the second in segmentation among 25 teams and 28 teams, respectively. This study corroborates that very deep CNNs with effective training mechanisms can be employed to solve complicated medical image analysis tasks, even with limited training data.

While ASCII art is a worldwide popular art form, automatic generating structure-based ASCII art from natural photographs remains challenging. The major challenge lies on extracting the perception-sensitive structure from the natural photographs so that a more concise ASCII art reproduction can be produced based on the structure. However, due to excessive amount of texture in natural photos, extracting perception-sensitive structure is not easy, especially when the structure may be weak and within the texture region. Besides, to fit different target text resolutions, the amount of the extracted structure should also be controllable. To tackle these challenges, we introduce a visual perception mechanism of non-classical receptive field modulation (non-CRF modulation) from physiological findings to this ASCII art application, and propose a new model of non-CRF modulation which can better separate the weak structure from the crowded texture, and also better control the scale of texture suppression. Thanks to our non-CRF model, more sensible ASCII art reproduction can be obtained. In addition, to produce more visually appealing ASCII arts, we propose a novel optimization scheme to obtain the optimal placement of proportional-font characters. We apply our method on a rich variety of images, and visually appealing ASCII art can be obtained in all cases.

The number of mitoses per tissue area gives an important aggressiveness indication of the invasive breast carcinoma.However, automatic mitosis detection in histology images remains a challenging problem. Traditional methods either employ hand-crafted features to discriminate mitoses from other cells or construct a pixel-wise classifier to label every pixel in a sliding window way. While the former suffers from the large shape variation of mitoses and the existence of many mimics with similar appearance, the slow speed of the later prohibits its use in clinical practice.In order to overcome these shortcomings, we propose a fast and accurate method to detect mitosis by designing a novel deep cascaded convolutional neural network, which is composed of two components. First, by leveraging the fully convolutional neural network, we propose a coarse retrieval model to identify and locate the candidates of mitosis while preserving a high sensitivity.Based on these candidates, a fine discrimination model utilizing knowledge transferred from cross-domain is developed to further single out mitoses from hard mimics.Our approach outperformed other methods by a large margin in 2014 ICPR MITOS-ATYPIA challenge in terms of detection accuracy. When compared with the state-of-the-art methods on the 2012 ICPR MITOSIS data (a smaller and less challenging dataset), our method achieved comparable or better results with a roughly 60 times faster speed.

Abstract This paper presents a new image smoothing method that better preserves prominent structures. Our method is inspired by the recent non‐local image processing techniques on the patch grouping and filtering. Overall, it has three major contributions over previous works. First, we employ the diffusion map as the guidance image to improve the accuracy of patch similarity estimation using the region covariance descriptor. Second, we model structure‐preserving image smoothing as a low‐rank matrix recovery problem, aiming at effectively filtering the texture information in similar patches. Lastly, we devise an objective function, namely the weighted robust principle component analysis (WRPCA), by regularizing the low rank with the weighted nuclear norm and sparsity pursuit with L1norm, and solve this non‐convex WRPCA optimization problem by adopting the alternative direction method of multipliers (ADMM) technique. We experiment our method with a wide variety of images and compare it against several state‐of‐the‐art methods. The results show that our method achieves better structure preservation and texture suppression as compared to other methods. We also show the applicability of our method on several image processing tasks such as edge detection, texture enhancement and seam carving.

Cerebral microbleeds (CMBs) are small haemorrhages nearby blood vessels. They have been recognized as important diagnostic biomarkers for many cerebrovascular diseases and cognitive dysfunctions. In current clinical routine, CMBs are manually labelled by radiologists but this procedure is laborious, time-consuming, and error prone. In this paper, we propose a novel automatic method to detect CMBs from magnetic resonance (MR) images by exploiting the 3D convolutional neural network (CNN). Compared with previous methods that employed either low-level hand-crafted descriptors or 2D CNNs, our method can take full advantage of spatial contextual information in MR volumes to extract more representative high-level features for CMBs, and hence achieve a much better detection accuracy. To further improve the detection performance while reducing the computational cost, we propose a cascaded framework under 3D CNNs for the task of CMB detection. We first exploit a 3D fully convolutional network (FCN) strategy to retrieve the candidates with high probabilities of being CMBs, and then apply a well-trained 3D CNN discrimination model to distinguish CMBs from hard mimics. Compared with traditional sliding window strategy, the proposed 3D FCN strategy can remove massive redundant computations and dramatically speed up the detection process. We constructed a large dataset with 320 volumetric MR scans and performed extensive experiments to validate the proposed method, which achieved a high sensitivity of 93.16% with an average number of 2.74 false positives per subject, outperforming previous methods using low-level descriptors or 2D CNNs by a significant margin. The proposed method, in principle, can be adapted to other biomarker detection tasks from volumetric medical data.

Registration and fusion of magnetic resonance (MR) and 3D transrectal ultrasound (TRUS) images of the prostate gland can provide high-quality guidance for prostate interventions. However, accurate MR-TRUS registration remains a challenging task, due to the great intensity variation between two modalities, the lack of intrinsic fiducials within the prostate, the large gland deformation caused by the TRUS probe insertion, and distinctive biomechanical properties in patients and prostate zones. To address these challenges, a personalized model-to-surface registration approach is proposed in this study. The main contributions of this paper can be threefold. First, a new personalized statistical deformable model (PSDM) is proposed with the finite element analysis and the patient-specific tissue parameters measured from the ultrasound elastography. Second, a hybrid point matching method is developed by introducing the modality independent neighborhood descriptor (MIND) to weight the Euclidean distance between points to establish reliable surface point correspondence. Third, the hybrid point matching is further guided by the PSDM for more physically plausible deformation estimation. Eighteen sets of patient data are included to test the efficacy of the proposed method. The experimental results demonstrate that our approach provides more accurate and robust MR-TRUS registration than state-of-the-art methods do. The averaged target registration error is 1.44 mm, which meets the clinical requirement of 1.9 mm for the accurate tumor volume detection. It can be concluded that the presented method can effectively fuse the heterogeneous image information in the elastography, MR, and TRUS to attain satisfactory image alignment performance.

Automatic localization of the standard plane containing complicated anatomical structures in ultrasound (US) videos remains a challenging problem. In this paper, we present a learning-based approach to locate the fetal abdominal standard plane (FASP) in US videos by constructing a domain transferred deep convolutional neural network (CNN). Compared with previous works based on low-level features, our approach is able to represent the complicated appearance of the FASP and hence achieve better classification performance. More importantly, in order to reduce the overfitting problem caused by the small amount of training samples, we propose a transfer learning strategy, which transfers the knowledge in the low layers of a base CNN trained from a large database of natural images to our task-specific CNN. Extensive experiments demonstrate that our approach outperforms the state-of-the-art method for the FASP localization as well as the CNN only trained on the limited US training samples. The proposed approach can be easily extended to other similar medical image computing problems, which often suffer from the insufficient training samples when exploiting the deep CNN to represent high-level features.

Most mesh denoising techniques utilize only either the facet normal field or the vertex normal field of a mesh surface. The two normal fields, though contain some redundant geometry information of the same model, can provide additional information that the other field lacks. Thus, considering only one normal field is likely to overlook some geometric features. In this paper, we take advantage of the piecewise consistent property of the two normal fields and propose an effective framework in which they are filtered and integrated using a novel method to guide the denoising process. Our key observation is that, decomposing the inconsistent field at challenging regions into multiple piecewise consistent fields makes the two fields complementary to each other and produces better results. Our approach consists of three steps: vertex classification, bi-normal filtering, and vertex position update. The classification step allows us to filter the two fields on a piecewise smooth surface rather than a surface that is smooth everywhere. Based on the piecewise consistence of the two normal fields, we filtered them using a piecewise smooth region clustering strategy. To benefit from the bi-normal filtering, we design a quadratic optimization algorithm for vertex position update. Experimental results on synthetic and real data show that our algorithm achieves higher quality results than current approaches on surfaces with multifarious geometric features and irregular surface sampling.

This paper presents a novel approach to simulating turbulent flows by developing an adaptive multirelaxation scheme in the framework of lattice Boltzmann equation (LBE). Existing LBE methods in graphics simulations are usually insufficient for turbulent flows since the collision term disturbs the underlying stability and accuracy. We adopt LBE with the multiple relaxation time (MRT) collision model (MRT-LBE), and address this issue by enhancing the collision-term modeling. First, we employ renormalization group analysis and formulate a new turbulence model with an adaptive correction method to compute more appropriate eddy viscosities on a uniform lattice structure. Efficient algebraic calculations are retained with small-scale turbulence details while maintaining the system stability. Second, we note that for MRT-LBE, predicting single eddy viscosity per lattice node may still result in instability. Hence, we simultaneously predict multiple eddy viscosities for stress-tensor-related elements, thereby asynchronously computing multiple relaxation parameters to further enhance the MRT-LBE stability. With these two new strategies, turbulent flows can be simulated with finer visual details even on coarse grid configurations. We demonstrate our results by simulating and visualizing various turbulent flows, particularly with smoke animations, where stable turbulent flows with high Reynolds numbers can be faithfully produced.

Abstract State‐of‐theart normal filters usually denoise each face normal using its entire anisotropic neighborhood. However, enforcing these filters indiscriminately on the anisotropic neighborhood will lead to feature blurring, especially in challenging regions with shallow features. We develop a novel mesh denoising framework which can effectively preserve features with various sizes. Our idea is inspired by the observation that the underlying surface of a noisy mesh is piecewise smooth. In this regard, it is more desirable that we denoise each face normal within its piecewise smooth region (we call such a region as an isotropic subneighborhood) instead of using the anisotropic neighborhood. To achieve this, we first classify mesh faces into several types using a face normal tensor voting and then perform a normal filter to obtain a denoised coarse normal field. Based on the results of normal classification and the denoised coarse normal field, we segment the anisotropic neighborhood of every feature face into a number of isotropic subneighborhoods via local spectral clustering. Thus face normal filtering can be performed again on the isotropic subneighborhoods and produce a more accurate normal field. Extensive tests on various models demonstrate that our method can achieve better performance than state‐of‐theart normal filters, especially in challenging regions with features.

Ultrasound-guided needle placement is a key step in a lot of radiological intervention procedures such as biopsy, local anesthesia and fluid drainage. To help training future intervention radiologists, we develop a serious game to teach the skills involved. We introduce novel techniques for realistic simulation and integrate game elements for active and effective learning. This game is designed in the context of needle placement training based on the some essential characteristics of serious games. Training scenarios are interactively generated via a block-based construction scheme. A novel example-based texture synthesis technique is proposed to simulate corresponding ultrasound images. Game levels are defined based on the difficulties of the generated scenarios. Interactive recommendation of desirable insertion paths is provided during the training as an adaptation mechanism. We also develop a fast physics-based approach to reproduce the shadowing effect of needles in ultrasound images. Game elements such as time-attack tasks, hints and performance evaluation tools are also integrated in our system. Extensive experiments are performed to validate its feasibility for training.

Realistic facial animation is important in many graphics applications, like animated feature films and computer games, to enrich human computer interaction. In this paper, we propose a physically-based facial animation approach employing knowledge from the anatomy and biomechanics of human facial muscles. First, the 3D face mesh is generated automatically by commercial software and the facial skin is represented by a nonlinear mass-spring system which simulates realistic elastic dynamics of human dermis. Then, a structure skull is attached to fit the face mesh. A set of anatomically consistent facial muscles are incorporated to model the forces deforming the face mesh. Finally, we extend Waters' muscle model to improve the combination of multiple muscle actions and to generate realistic expression. Experiments show that our method is superior to the traditional geometric model and can achieve comparative results with the commercial software FaceGen.