3D Human Pose Estimation (3D HPE) is vital in various applications, from person re-identification and action recognition to virtual reality. However, the reliance on annotated 3D data collected in controlled environments poses challenges for generalization to diverse in-the-wild scenarios. Existing domain adaptation (DA) paradigms like general DA and source-free DA for 3D HPE overlook the issues of non-stationary target pose datasets. To address these challenges, we propose a novel task named lifelong domain adaptive 3D HPE. To our knowledge, we are the first to introduce the lifelong domain adaptation to the 3D HPE task. In this lifelong DA setting, the pose estimator is pretrained on the source domain and subsequently adapted to distinct target domains. Moreover, during adaptation to the current target domain, the pose estimator cannot access the source and all the previous target domains. The lifelong DA for 3D HPE involves overcoming challenges in adapting to current domain poses and preserving knowledge from previous domains, particularly combating catastrophic forgetting. We present an innovative Generative Adversarial Network (GAN) framework, which incorporates 3D pose generators, a 2D pose discriminator, and a 3D pose estimator. This framework effectively mitigates domain shifts and aligns original and augmented poses. Moreover, we construct a novel 3D pose generator paradigm, integrating pose-aware, temporal-aware, and domain-aware knowledge to enhance the current domain's adaptation and alleviate catastrophic forgetting on previous domains. Our method demonstrates superior performance through extensive experiments on diverse domain adaptive 3D HPE datasets.

Majority-Inverter Graph (MIG) emerges as a promising data structure for logic optimization and synthesis, offering a more compact representation for logic functions compared to traditional AND/OR-Inverter graphs. Consequently, the MIG finds widespread application in digital circuit design, particularly in quantum circuits and superconducting adiabatic quantum-flux-parametron logic circuits. Currently, logic optimization techniques for MIG mainly fall into two categories: (i) logic rewriting with predefined more compact sub-structures and (ii) logic resubstitution with already existing logic in the Boolean network. However, the inherent complexity of MIG logic and the limitation imposed by the input scale of sub-structures significantly impact the performance of these methods. To address these challenges, this paper proposes eLogic, a novel depth-oriented MIG logic rewriting framework using e-graphs, to minimize the depth and size of MIG. The eLogic utilizes the e-graphs, a data structure for efficient computation with equalities between terms, to minimize the depth and size of the cone delimited by the cut. The experimental results on the EPFL benchmark demonstrate the effectiveness of eLogic. It is noteworthy that eLogic is open-sourced on https://github.com/Flians/eLogic.

Abstract Word-level verification of arithmetic circuits with large operands typically relies on arbitrary-precision arithmetic, which can lead to significant computational overhead as word sizes grow. In this paper, we present a hybrid algebraic verification technique based on polynomial reasoning that combines linear and nonlinear rewriting. Our approach relies on multimodular reasoning using homomorphic images, where computations are performed in parallel modulo different primes, thereby avoiding any large-integer arithmetic. We implement the proposed method in the verification tool TalisMan2.0 and evaluate it on a suite of multiplier benchmarks. Our results show that hybrid multimodular reasoning significantly improves upon existing approaches.

We propose PoseGaussian, a pose-guided Gaussian Splatting framework for high-fidelity human novel view synthesis. Human body pose serves a dual purpose in our design: as a structural prior, it is fused with a color encoder to refine depth estimation; as a temporal cue, it is processed by a dedicated pose encoder to enhance temporal consistency across frames. These components are integrated into a fully differentiable, end-to-end trainable pipeline. Unlike prior works that use pose only as a condition or for warping, PoseGaussian embeds pose signals into both geometric and temporal stages to improve robustness and generalization. It is specifically designed to address challenges inherent in dynamic human scenes, such as articulated motion and severe self-occlusion. Notably, our framework achieves real-time rendering at 100 FPS, maintaining the efficiency of standard Gaussian Splatting pipelines. We validate our approach on ZJU-MoCap, THuman2.0, and in-house datasets, demonstrating state-of-the-art performance in perceptual quality and structural accuracy (PSNR 30.86, SSIM 0.979, LPIPS 0.028).

Videos contain multi-modal content, and exploring multi-branch cross-modal interactions with natural language queries can be of benefit to the text-video retrieval task (TVR). However, recent methods applying the large-scale pre-trained CLIP model for TVR only focus on visual cues in videos. Furthermore, traditional methods of simply concatenating multimodal features do not exploit fine-grained cross-modal information in videos. In this paper, we propose a multi-branch multi-modal hybrid fusion (M2HF) network to hierarchically explore interaction between text queries and other modality content in videos. Specifically, M2HF first fuses visual features extracted by CLIP with audio and motion features extracted from videos to obtain fused audio-visual features and motion-visual features respectively. The multi-modal completion problem is also considered and solved in this process. Then, visual features, audio-visual features, motion-visual features, and text extracted from the video are used to establish cross-modal relationships with caption text queries using a multibranch approach. The retrieval outputs from all branches are then fused to obtain the final text-video retrieval results. Our framework provides two kinds of training strategies, using an ensemble approach and an end-to-end approach. Moreover, a novel multi-modal loss function is proposed to balance the contributions of each modality for efficient end-to-end training. M2HF allows us to obtain state-of-the-art results on various benchmarks: Rank@1 of 66.0%, 68.6%, 33.9%, 57.4%, and 57.3% on MSR-VTT, MSVD, LSMDC, DiDeMo, and ActivityNet, respectively.

Human mesh recovery (HMR) is crucial in many computer vision applications; from health to arts and entertainment. HMR from monocular images has predominantly been addressed by deterministic methods that output a single prediction for a given 2D image. However, HMR from a single image is an ill-posed problem due to depth ambiguity and occlusions. Probabilistic methods have attempted to address this by generating and fusing multiple plausible 3D reconstructions, but their performance has often lagged behind deterministic approaches. In this paper, we introduce GenHMR, a novel generative framework that reformulates monocular HMR as an image-conditioned generative task, explicitly modeling and mitigating uncertainties in the 2D-to-3D mapping process. GenHMR comprises two key components: (1) a pose tokenizer to convert 3D human poses into a sequence of discrete tokens in a latent space, and (2) an image-conditional masked transformer to learn the probabilistic distributions of the pose tokens, conditioned on the input image prompt along with randomly masked token sequence. During inference, the model samples from the learned conditional distribution to iteratively decode high-confidence pose tokens, thereby reducing 3D reconstruction uncertainties. To further refine the reconstruction, a 2D pose-guided refinement technique is proposed to directly fine-tune the decoded pose tokens in the latent space, which forces the projected 3D body mesh to align with the 2D pose clues. Experiments on benchmark datasets demonstrate that GenHMR significantly outperforms state-of-the-art methods.

We introduce Motion-Grounded Video Reasoning, a new motion understanding task that requires generating visual answers (video segmentation masks) according to the input question, and hence needs implicit spatiotemporal reasoning and grounding. It extends existing spatiotemporal grounding work focusing on explicit action/motion grounding, to a more general format by enabling implicit reasoning via questions. To facilitate the development of the new task, we collect a large-scale dataset called GroundMoRe, which comprises 1,715 video clips, 249K object masks that are deliberately designed with 4 question types for benchmarking deep and comprehensive motion reasoning abilities. GroundMoRe uniquely requires models to generate visual answers, providing a more concrete and visually interpretable response than plain texts. It evaluates models on both spatiotemporal grounding and reasoning, fostering to address complex challenges in motion-related video reasoning, temporal perception, and pixel-level understanding. Furthermore, we introduce a novel baseline model named MoRA, which achieves respectable performance on GroundMoRe outperforming the best existing visual grounding baseline model by an average of 21.5% relatively. We hope this novel and challenging task will pave the way for future advancements in robust and general motion understanding via video reasoning segmentation. Project available at: https://groundmore.github.io/

Temporal awareness is essential for video large language models (LLMs) to understand and reason about events within long videos, enabling applications like dense video captioning and temporal video grounding in a unified system. However, the scarcity of long videos with detailed captions and precise temporal annotations limits their temporal awareness. In this paper, we propose Seq2Time, a data-oriented training paradigm that leverages sequences of images and short video clips to enhance temporal awareness in long videos. By converting sequence positions into temporal annotations, we transform large-scale image and clip captioning datasets into sequences that mimic the temporal structure of long videos, enabling self-supervised training with abundant time-sensitive data. To enable sequence-to-time knowledge transfer, we introduce a novel time representation that unifies positional information across image sequences, clip sequences, and long videos. Experiments demonstrate the effectiveness of our method, achieving a 27.6% improvement in F1 score and 44.8% in CIDEr on the YouCook2 benchmark and a 14.7% increase in recall on the Charades-STA benchmark compared to the baseline. Project available at: https://seq2time.github.io/

Visual Prompt Tuning (VPT) has become a promising solution for Parameter-Efficient Fine-Tuning (PEFT) approach for Vision Transformer (ViT) models by partially fine-tuning learnable tokens while keeping most model parameters frozen. Recent research has explored modifying the connection structures of the prompts. However, the fundamental correlation and distribution between the prompts and image tokens remain unexplored. In this paper, we leverage metric learning techniques to investigate how the distribution of prompts affects fine-tuning performance. Specifically, we propose a novel framework, Distribution Aware Visual Prompt Tuning (DA-VPT), to guide the distributions of the prompts by learning the distance metric from their class-related semantic data. Our method demonstrates that the prompts can serve as an effective bridge to share semantic information between image patches and the class token. We extensively evaluated our approach on popular benchmarks in both recognition and segmentation tasks. The results demonstrate that our approach enables more effective and efficient fine-tuning of ViT models by leveraging semantic information to guide the learning of the prompts, leading to improved performance on various downstream vision tasks. The code is released on https://github.com/Noahsark/DA-VPT.

In this paper, we propose a dual-condition diffusion pre-training model named SignDIFF that can generate human sign language speakers from a skeleton pose. SignDiff has a novel Frame Reinforcement Network called FR-Net, similar to dense human pose estimation work, which enhances the correspondence between text lexical symbols and sign language dense pose frames, reduces the occurrence of multiple fingers in the diffusion model. In addition, we propose a new method for American Sign Language Production (ASLP), which can generate ASL skeletal pose videos from text input, integrating two new improved modules and a new loss function to improve the accuracy and quality of sign language skeletal posture and enhance the ability of the model to train on largescale data. We propose a simple baseline for ASL production and report the scores of 17.19 and 12.85 on BLEU-4 on the How2Sign dev/test sets. We evaluated our model on the previous mainstream dataset PHOENIX14T, and our method achieved the SOTA results. In addition, our image quality far exceeds all previous results by 10 percentage points in terms of SSIM.

Human pose estimation (HPE) has received increasing attention recently due to its wide application in motion analysis, virtual reality, healthcare, etc. However, it suffers from the lack of labeled diverse real-world datasets due to the timeand labor-intensive annotation. To cope with the label deficiency issue, one common solution is to train the HPE models with easily available synthetic datasets (source) and apply them to real-world data (target) through domain adaptation (DA). Unfortunately, prevailing domain adaptation techniques within the HPE domain remain predominantly fixated on effecting alignment and aggregation between source and target features, often sidestepping the crucial task of excluding domain-specific representations. To rectify this, we introduce a novel framework that capitalizes on both representation aggregation and segregation for domain adaptive human pose estimation. Within this framework, we address the network architecture aspect by disentangling representations into distinct domain-invariant and domain-specific components, facilitating aggregation of domaininvariant features while simultaneously segregating domainspecific ones. Moreover, we tackle the discrepancy measurement facet by delving into various keypoint relationships and applying separate aggregation or segregation mechanisms to enhance alignment. Extensive experiments on various benchmarks, e.g., Human3.6M, LSP, H3D, and FreiHand, show that our method consistently achieves state-of-the-art performance.

Transformer-based human skeleton action recognition has been developed for years. However, the complexity and high parameter count demands of these models hinder their practical applications, especially in resource-constrained environments. In this work, we propose FreqMixForemrV2, which was built upon the Frequency-aware Mixed Transformer (FreqMixFormer) [17] for identifying subtle and discriminative actions with pioneered frequency-domain analysis. We design a lightweight architecture that maintains robust performance while significantly reducing the model complexity. This is achieved through a redesigned frequency operator that optimizes high-frequency and low-frequency parameter adjustments, and a simplified frequency-aware attention module. These improvements result in a substantial reduction in model parameters, enabling efficient deployment with only a minimal sacrifice in accuracy. Comprehensive evaluations of standard datasets (NTU RGB+D, NTU RGB+D 120, and NW-UCLA datasets) demonstrate that the proposed model achieves a superior balance between efficiency and accuracy, outperforming state-of-the-art methods with only $60 \%$ of the parameters. Our project is publicly available at: https://github.com/wenhanwu95/FreqMixFormer.

The role of high-quality pseudo-labels is pivotal in semi-supervised instance segmentation (SSIS). However, existing SSIS frameworks predominantly produce pseudo-labels at a single resolution, which can introduce noise that adversely affects the quality of learning at both the pixel level and in terms of class discrimination. This paper introduces the Dual-Resolution Pseudo-Label Integration for Semi-Supervised Instance Segmentation (DuPI), a novel framework designed to enhance learning by integrating pseudo-labels derived from dual-resolution inputs. The DuPI framework incorporates a Dual-Resolution Pseudo-Label Correction (DPC) module, which refines pseudo-labels through a process of cross-resolution rectification and fusion. Furthermore, the framework introduces an Area-Adaptive Learning (AAL) strategy aimed at enhancing the quality of pseudo-labels sourced from extra-resolution inputs. The AAL strategy addresses the training challenges associated with small objects at lower resolutions by re-weighting pseudo-labels corresponding to tiny mask areas using Intersection over Union (IoU) metrics from the assignments. Experiments on the COCO and BDD100K datasets demonstrate that DuPI achieves state-of-the-art SSIS performance under various semi-supervised settings.

In this paper, we propose SimMLM, a simple yet powerful framework for multimodal learning with missing modalities. Unlike existing approaches that rely on sophisticated network architectures or complex data imputation techniques, SimMLM provides a generic and effective solution that can adapt to various missing modality scenarios with improved accuracy and robustness. Specifically, SimMLM consists of a generic Dynamic Mixture of Modality Experts (DMoME) architecture, featuring a dynamic, learnable gating mechanism that automatically adjusts each modality's contribution in both full and partial modality settings. A key innovation of SimMLM is the proposed More vs. Fewer (MoFe) ranking loss, which ensures that task accuracy improves or remains stable as more modalities are made available. This aligns the model with an intuitive principle: removing one or more modalities should not increase accuracy. We validate SimMLM on multimodal medical image segmentation (BraTS 2018) and multimodal classification (UPMC Food-101, avMNIST) tasks, where it consistently surpasses competitive methods, demonstrating superior accuracy, interpretability, robustness, and reliability across both complete and missing modality scenarios at test time.

Due to the challenges of manually collecting accurate editing data, existing datasets are typically constructed using various automated methods, leading to noisy supervision signals caused by the mismatch between editing instructions and original-edited image pairs. Recent efforts attempt to improve editing models through generating higher-quality edited images, pre-training on recognition tasks, or introducing vision-language models (VLMs) but fail to resolve this fundamental issue. In this paper, we offer a novel solution by constructing more effective editing instructions for given image pairs. This includes rectifying the editing instructions to better align with the original-edited image pairs and using contrastive editing instructions to further enhance their effectiveness. Specifically, we find that editing models exhibit specific generation attributes at different inference steps, independent of the text. Based on these prior attributes, we define a unified guide for VLMs to rectify editing instructions. However, there are some challenging editing scenarios that cannot be resolved solely with rectified instructions. To this end, we further construct contrastive supervision signals with positive and negative instructions and introduce them into the model training using triplet loss, thereby further facilitating supervision effectiveness. Our method does not require the VLM modules or pre-training tasks used in previous work, offering a more direct and efficient way to provide better supervision signals, and providing a novel, simple, and effective solution for instruction-based image editing. Results on multiple benchmarks demonstrate that our method significantly outperforms existing approaches. Compared with previous SOTA SmartEdit, we achieve 9.19% improvements on the Real-Edit benchmark with 30x less training data and 13x smaller model size.

Zero-shot skeleton-based action recognition aims to develop models capable of identifying actions beyond the categories encountered during training. Previous approaches have primarily focused on aligning visual and semantic representations but often overlooked the importance of fine-grained action patterns in the semantic space (e.g., the hand movements in drinking water and brushing teeth). To address these limitations, we propose a Frequency-Semantic Enhanced Variational Autoencoder (FS-VAE) to explore the skeleton semantic representation learning with frequency decomposition. FS-VAE consists of three key components: 1) a frequency-based enhancement module with high- and low-frequency adjustments to enrich the skeletal semantics learning and improve the robustness of zero-shot action recognition; 2) a semantic-based action description with multilevel alignment to capture both local details and global correspondence, effectively bridging the semantic gap and compensating for the inherent loss of information in skeleton sequences; 3) a calibrated cross-alignment loss that enables valid skeleton-text pairs to counterbalance ambiguous ones, mitigating discrepancies and ambiguities in skeleton and text features, thereby ensuring robust alignment. Evaluations on the benchmarks demonstrate the effectiveness of our approach, validating that frequency-enhanced semantic features enable robust differentiation of visually and semantically similar action clusters, improving zero-shot action recognition.

Federated prompt learning benefits federated learning with CLIP-like Vision-Language Model's (VLM's) robust representation learning ability through prompt learning. However, current federated prompt learning methods are habitually restricted to the traditional FL paradigm, where the participating clients are generally only allowed to download a single globally aggregated model from the server. While justifiable for training full-sized models under federated settings, in this work, we argue that this paradigm is ill-suited for lightweight prompts. By facilitating the clients to download multiple pre-aggregated prompts as fixed non-local experts, we propose Personalized Federated Mixture of Adaptive Prompts (pFedMoAP), a novel FL framework that personalizes the prompt learning process through the lens of Mixture of Experts (MoE). pFedMoAP implements a local attention-based gating network that learns to generate enhanced text features for better alignment with local image data, benefiting from both local and downloaded non-local adaptive prompt experts. Extensive experiments on 9 datasets under various federated settings demonstrate the efficacy of the proposed pFedMoAP algorithm. The code is available at https://github.com/ljaiverson/pFedMoAP.

Recent advancements in 3D reconstruction methods and vision-language models have propelled the development of multi-modal 3D scene understanding, which has vital applications in robotics, autonomous driving, and virtual/augmented reality. However, current multi-modal scene understanding approaches have naively embedded semantic representations into 3D reconstruction methods without striking a balance between visual and language modalities, which leads to unsatisfying semantic rasterization of translucent or reflective objects, as well as over-fitting on color modality. To alleviate these limitations, we propose a solution that adequately handles the distinct visual and semantic modalities, i.e., a 3D vision-language Gaussian splatting model for scene understanding, to put emphasis on the representation learning of language modality. We propose a novel cross-modal rasterizer, using modality fusion along with a smoothed semantic indicator for enhancing semantic rasterization. We also employ a camera-view blending technique to improve semantic consistency between existing and synthesized views, thereby effectively mitigating over-fitting. Extensive experiments demonstrate that our method achieves state-of-the-art performance in open-vocabulary semantic segmentation, surpassing existing methods by a significant margin.

Autonomous driving systems have made significant advances in Q&A, perception, prediction, and planning based on local visual information, yet they struggle to incorporate broader navigational context that human drivers routinely utilize. We address this critical gap between local sensor data and global navigation information by proposing NavigScene, an auxiliary navigation-guided natural language dataset that simulates a human-like driving environment within autonomous driving systems. Moreover, we develop three complementary paradigms to leverage NavigScene: (1) Navigation-guided Reasoning, which enhances vision-language models by incorporating navigation context into the prompting approach; (2) Navigation-guided Preference Optimization, a reinforcement learning method that extends Direct Preference Optimization to improve vision-language model responses by establishing preferences for navigation-relevant summarized information; and (3) Navigation-guided Vision-Language-Action model, which integrates navigation guidance and vision-language models with conventional driving models through feature fusion. Extensive experiments demonstrate that our approaches significantly improve performance across perception, prediction, planning, and question-answering tasks by enabling reasoning capabilities beyond visual range and improving generalization to diverse driving scenarios. This work represents a significant step toward more comprehensive autonomous driving systems capable of navigating complex, unfamiliar environments with greater reliability and safety.

To enhance controllability in text-to-image generation, ControlNet introduces image-based control signals, while ControlNet++ improves pixel-level cycle consistency between generated images and the input control signal. To avoid the prohibitive cost of back-propagating through the sampling process, ControlNet++ optimizes only low-noise timesteps (e.g., $t < 200$) using a single-step approximation, which not only ignores the contribution of high-noise timesteps but also introduces additional approximation errors. A straightforward alternative for optimizing controllability across all timesteps is Direct Preference Optimization (DPO), a fine-tuning method that increases model preference for more controllable images ($I^{w}$) over less controllable ones ($I^{l}$). However, due to uncertainty in generative models, it is difficult to ensure that win--lose image pairs differ only in controllability while keeping other factors, such as image quality, fixed. To address this, we propose performing preference learning over control conditions rather than generated images. Specifically, we construct winning and losing control signals, $\mathbf{c}^{w}$ and $\mathbf{c}^{l}$, and train the model to prefer $\mathbf{c}^{w}$. This method, which we term \textit{Condition Preference Optimization} (CPO), eliminates confounding factors and yields a low-variance training objective. Our approach theoretically exhibits lower contrastive loss variance than DPO and empirically achieves superior results. Moreover, CPO requires less computation and storage for dataset curation. Extensive experiments show that CPO significantly improves controllability over the state-of-the-art ControlNet++ across multiple control types: over $10\%$ error rate reduction in segmentation, $70$--$80\%$ in human pose, and consistent $2$--$5\%$ reductions in edge and depth maps. The error rate is defined as the difference between the evaluated controllability and the oracle results. Our project is available \textcolor{blue}{\href{https://zonglinl.github.io/CPO_page}{here}}.

Novel view synthesis (NVS) is crucial in computer vision and graphics, with wide applications in AR, VR, and autonomous driving. While 3D Gaussian Splatting (3DGS) enables real-time rendering with high appearance fidelity, it suffers from multi-view inconsistencies, limiting geometric accuracy. In contrast, 2D Gaussian Splatting (2DGS) enforces multi-view consistency but compromises texture details. To address these limitations, we propose Exchangeable Gaussian Splatting (EGGS), a hybrid representation that integrates 2D and 3D Gaussians to balance appearance and geometry. To achieve this, we introduce Hybrid Gaussian Rasterization for unified rendering, Adaptive Type Exchange for dynamic adaptation between 2D and 3D Gaussians, and Frequency-Decoupled Optimization that effectively exploits the strengths of each type of Gaussian representation. Our CUDA-accelerated implementation ensures efficient training and inference. Extensive experiments demonstrate that EGGS outperforms existing methods in rendering quality, geometric accuracy, and efficiency, providing a practical solution for high-quality NVS.

Aerial imagery analysis is critical for many research fields. However, obtaining frequent high-quality aerial images is not always accessible due to its high effort and cost requirements. One solution is to use the Ground-to-Aerial (G2A) technique to synthesize aerial images from easily collectible ground images. However, G2A is rarely studied, because of its challenges, including but not limited to, the drastic view changes, occlusion, and range of visibil-ity. In this paper, we present a novel Geometric Preserving Ground-to-Aerial (G2A) image synthesis (GPG2A) model that can generate realistic aerial images from ground images. GPG2A consists of two stages. The first stage predicts the Bird's Eye View (BEV) segmentation (referred to as the BEV layout map) from the ground image. The second stage synthesizes the aerial image from the predicted BEV layout map and text descriptions of the ground image. To train our model, we present a new multimodal cross-view dataset, namely VIGORv2, built upon VIGOR [64] with newly collected aerial images, maps, and text descriptions. Our extensive experiments illustrate that GPG2A synthesizes better geometry-preserved aerial images than existing models. We also present two applications, data augmentation for cross-view geo-localization and sketch-based region search, to further verify the effectiveness of our GPG2A. The code and dataset are available at https://github.com/AhmadArrabi/GPG2A.

Federated learning (FL) enables multiple clients to train models collectively while preserving data privacy. However, FL faces challenges in terms of communication cost and data heterogeneity. One-shot federated learning has emerged as a solution by reducing communication rounds, improving efficiency, and providing better security against eavesdropping attacks. Nevertheless, data heterogeneity remains a significant challenge, impacting performance. This work explores the effectiveness of diffusion models in oneshot FL, demonstrating their applicability in addressing data heterogeneity and improving FL performance. Additionally, we investigate the utility of our diffusion model approach, FedDiff, compared to other one-shot FL methods under differential privacy (DP). Furthermore, to improve generated sample quality under DP settings, we propose a pragmatic Fourier Magnitude Filtering (FMF) method, enhancing the effectiveness of the generated data for global model training. Code available at https://github.com/mmendiet/FedDiff.

Human mesh recovery (HMR) provides rich human body information for various real-world applications such as gaming, human-computer interaction, and virtual reality. While image-based HMR methods have achieved impressive results, they often struggle to recover humans in dynamic scenarios, leading to temporal inconsistencies and non-smooth 3D motion predictions due to the absence of human motion. In contrast, video-based approaches leverage temporal information to mitigate this issue. In this paper, we present DiffMesh, an innovative motion-aware diffusion framework for video-based HMR. DiffMesh establishes a bridge between diffusion models and human motion, efficiently generating accurate and smooth output mesh sequences by incorporating human motion within the forward process and reverse process in the diffusion model. Extensive experiments are conducted on the widely used datasets (Human3.6M [15] and 3DPW [48]), which demonstrate the effectiveness and efficiency of our DiffMesh. Visual comparisons in real-world scenarios further highlight DiffMesh's suitability for practical applications. The project webpage is: https://zczcwh.github.io/diffmesh_page/

Deep Metric Learning (DML) plays an important role in modern computer vision research, where we learn a distance metric for a set of image representations. Recent DML techniques utilize the proxy to interact with the corresponding image samples in the embedding space. However, existing proxy-based DML methods focus on learning individual proxy-to-sample distance, while the overall distribution of samples and proxies lacks attention. In this paper, we present a novel proxy-based DML framework that focuses on aligning the sample and proxy distributions to improve the efficiency of proxy-based DML losses. Specifically, we propose the Data-Augmented Domain Adaptation (DADA) method to adapt the domain gap between the group of samples and proxies. To the best of our knowledge, we are the first to leverage domain adaptation to boost the performance of proxy-based DML. We show that our method can be easily plugged into existing proxy-based DML losses. Our experiments on benchmarks, including the popular CUB-200-2011, CARS196, Stanford Online Products, and In-Shop Clothes Retrieval, show that our learning algorithm significantly improves the existing proxy losses and achieves superior results compared to the existing methods. The code and Appendix are available at: https://github.com/Noahsark/DADA

The past few years has witnessed the dominance of Graph Convolutional Networks (GCNs) over human motion prediction. Various styles of graph convolutions have been proposed, with each one meticulously designed and incorporated into a carefully-crafted network architecture. This paper breaks the limits of existing knowledge by proposing Universal Graph Convolution (UniGC), a novel graph convolution concept that re-conceptualizes different graph convolutions as its special cases. Leveraging UniGC on network-level, we propose GCNext, a novel GCN-building paradigm that dynamically determines the best-fitting graph convolutions both sample-wise and layer-wise. GCNext offers multiple use cases, including training a new GCN from scratch or refining a preexisting GCN. Experiments on Human3.6M, AMASS, and 3DPW datasets show that, by incorporating unique module-to-network designs, GCNext yields up to 9x lower computational cost than existing GCN methods, on top of achieving state-of-the-art performance. Our code is available at https://github.com/BradleyWang0416/GCNext.

The Few-Shot Segmentation (FSS) aims to accomplish the novel class segmentation task with a few annotated images. Current FSS research based on meta-learning focuses on designing a complex interaction mechanism between the query and support feature. However, unlike humans who can rapidly learn new things from limited samples, the existing approach relies solely on fixed feature matching to tackle new tasks, lacking adaptability. In this paper, we propose a novel framework based on the adapter mechanism, namely Adaptive FSS, which can efficiently adapt the existing FSS model to the novel classes. In detail, we design the Prototype Adaptive Module (PAM), which utilizes accurate category information provided by the support set to derive class prototypes, enhancing class-specific information in the multi-stage representation. In addition, our approach is compatible with diverse FSS methods with different backbones by simply inserting PAM between the layers of the encoder. Experiments demonstrate that our method effectively improves the performance of the FSS models (e.g., MSANet, HDMNet, FPTrans, and DCAMA) and achieves new state-of-the-art (SOTA) results (i.e., 72.4% and 79.1% mIoU on PASCAL-5i 1-shot and 5-shot settings, 52.7% and 60.0% mIoU on COCO-20i 1-shot and 5-shot settings). Our code is available at https://github.com/jingw193/AdaptiveFSS.

Federated learning (FL) has emerged as a promising paradigm for enabling the collaborative training of models without centralized access to the raw data on local devices. In the typical FL paradigm (e.g., FedAvg), model weights are sent to and from the server each round to participating clients. Recently, the use of small pre-trained models has been shown to be effective in federated learning optimization and improving convergence. However, recent state-of-the-art pre-trained models are getting more capable but also have more parameters, known as the "Foundation Models." In conventional FL, sharing the enormous model weights can quickly put a massive communication burden on the system, especially if more capable models are employed. Can we find a solution to enable those strong and readily available pre-trained models in FL to achieve excellent performance while simultaneously reducing the communication burden? To this end, we investigate the use of parameter-efficient fine-tuning in federated learning and thus introduce a new framework: FedPEFT. Specifically, we systemically evaluate the performance of FedPEFT across a variety of client stability, data distribution, and differential privacy settings. By only locally tuning and globally sharing a small portion of the model weights, significant reductions in the total communication overhead can be achieved while maintaining competitive or even better performance in a wide range of federated learning scenarios, providing insight into a new paradigm for practical and effective federated systems.

Due to the resource-intensive nature of training vision- language models on expansive video data, a majority of studies have centered on adapting pre-trained image- language models to the video domain. Dominant pipelines propose to tackle the visual discrepancies with additional temporal learners while overlooking the substantial discrepancy for web-scaled descriptive narratives and concise action category names, leading to less distinct semantic space and potential performance limitations. In this work, we prioritize the refinement of text knowledge to facilitate generalizable video recognition. To address the limitations of the less distinct semantic space of category names, we prompt a large language model (LLM) to augment action class names into Spatio-Temporal Descriptors thus bridging the textual discrepancy and serving as a knowledge base for general recognition. Moreover, to assign the best descriptors with different video instances, we propose Optimal Descriptor Solver, forming the video recognition problem as solving the optimal matching flow across frame-level representations and descriptors. Comprehensive evaluations in zero-shot, few-shot, and fully supervised video recognition highlight the effectiveness of our approach. Our best model achieves a state-of-the-art zero-shot accuracy of 75.1% on Kinetics-600.

We present Multi-View Attentive Contextualization (MvACon), a simple yet effective method for improving 2D-to-3D feature lifting in query-based multi-view 3D (MV3D) object detection. Despite remarkable progress witnessed in the field of query-based MV3D object detection, prior art often suffers from either the lack of exploiting high-resolution 2D features in dense attention-based lifting, due to high computational costs, or from insufficiently dense grounding of 3D queries to multi-scale 2D features in sparse attention-based lifting. Our proposed MvACon hits the two birds with one stone using a representationally dense yet computationally sparse attentive feature contextualization scheme that is agnostic to specific 2D-to-3D feature lifting approaches. In experiments, the proposed MvA-Con is thoroughly tested on the nuScenes benchmark, using both the BEVFormer and its recent 3D deformable attention (DFA3D) variant, as well as the PETR, showing consistent detection performance improvement, especially in enhancing performance in location, orientation, and velocity prediction. It is also tested on the Waymo-mini benchmark using BEVFormer with similar improvement. We qualitatively and quantitatively show that global cluster-based contexts effectively encode dense scene-level contexts for MV3D object detection. The promising results of our proposed MvA-Con reinforces the adage in computer vision - “(contextualized) feature matters”.

Recent advances in text-to-motion generation using dif-fusion and autoregressive models have shown promising re-sults. However, these models often suffer from a trade-off between real-time performance, high fidelity, and motion editability. To address this gap, we introduce MMM, a novel yet simple motion generation paradigm based on Masked Motion Model. MMM consists of two key components: (1) a motion tokenizer that transforms 3D human motion into a sequence of discrete tokens in latent space, and (2) a conditional masked motion transformer that learns to predict randomly masked motion tokens, conditioned on the pre-computed text tokens. By attending to motion and text to-kens in all directions, MMM explicitly captures inherent dependency among motion tokens and semantic mapping between motion and text tokens. During inference, this al-lows parallel and iterative decoding of multiple motion to-kens that are highly consistent with fine-grained text de-scriptions, therefore simultaneously achieving high-fidelity and high-speed motion generation. In addition, MMM has innate motion editability. By simply placing mask tokens in the place that needs editing, MMM automatically fills the gaps while guaranteeing smooth transitions between editing and non-editing parts. Extensive experiments on the HumanML3D and KIT-ML datasets demonstrate that MMM surpasses current leading methods in generating high-quality motion (evidenced by superior FID scores of 0.08 and 0.429), while offering advanced editing features such as body-part modification, motion in-betweening, and the synthesis of long motion sequences. In addition, MMM is two orders of magnitude faster on a single mid-range GPU than editable motion diffusion models. Our project page is available at https://exitudio.github.io/MMM-page/.

In-context learning provides a new perspective for multi-task modeling for vision and NLP. Under this setting, the model can perceive tasks from prompts and accomplish them without any extra task-specific head predictions or model fine-tuning. However, skeleton sequence modeling via in-context learning remains unexplored. Directly applying existing in-context models from other areas onto skeleton sequences fails due to the similarity between inter-frame and cross-task poses, which makes it exceptionally hard to perceive the task correctly from a subtle context. To address this challenge, we propose Skeleton-in-Context (SiC), an effective framework for in-context skeleton sequence modeling. Our SiC is able to handle multiple skeleton-based tasks simultaneously after a single training process and accomplish each task from context according to the given prompt. It can further generalize to new, unseen tasks according to customized prompts. To facilitate context perception, we additionally propose a task-unified prompt, which adaptively learns tasks of different natures, such as partial joint-level generation, sequence-level prediction, or 2D-to-3D motion prediction. We conduct extensive experiments to evaluate the effectiveness of our SiC on multiple tasks, including motion prediction, pose estimation, joint completion, and future pose estimation. We also evaluate its generalization capability on unseen tasks such as motion-in-between. These experiments show that our model achieves state-of-the-art multi-task performance and even outperforms single-task methods on certain tasks.

Vision Transformers (ViTs) have emerged as a promising solution to enable efficient 3D Human Mesh Recovery (HMR) in augmented and virtual reality (AR/VR) applications. Despite many advancements in algorithm design, it remains a challenge to efficiently accelerate ViT-based HMR due to high computational complexity, substantial memory footprint, and compromised data locality. In this paper, we propose VITA, a hardware and algorithm co-design framework for ViT-based HMR with improved performance and energy efficiency. Specifically, on the algorithm side, we propose an average pooling model to replace conventional multi-head attention, which is further optimized with improved data locality. On the hardware side, we propose an accelerator architecture that can efficiently support various dataflows and computations demanded by pooling, normalization, and convolution operations. We evaluate the proposed VITA, and the evaluation result shows that the proposed VITA design can achieve 5.05× and 69.12× speedups on average over the state-of-the-art GPUs and CPUs on HMR tasks.

Malaria is a major health issue worldwide, and its diagnosis requires scalable solutions that can work effectively with low-cost microscopes (LCM). Deep learning-based methods have shown success in computer-aided diagnosis from microscopic images. However, these methods need annotated images that show cells affected by malaria parasites and their life stages. Annotating images from LCM significantly increases the burden on medical experts compared to annotating images from high-cost microscopes (HCM). For this reason, a practical solution would be trained on HCM images which should generalize well on LCM images during testing. While earlier methods adopted a multi-stage learning process, they did not offer an end-to-end approach. In this work, we present an end-to-end learning framework, named CodaMal (COntrastive Domain Adpation for MALaria). In order to bridge the gap between HCM (training) and LCM (testing), we propose a domain adaptive contrastive loss. It reduces the domain shift by promoting similarity between the representations of HCM and its corresponding LCM image, without imposing an additional annotation burden. In addition, the training objective includes object detection objectives with carefully designed augmentations, ensuring the accurate detection of malaria parasites. On the publicly available large-scale M5-dataset, our proposed method shows a significant improvement of $\mathbf{1 6 \%}$ over the state-of-the-art methods in terms of the mean average precision metric (mAP), provides $21 \times$ speed improvement during inference and requires only half of the learnable parameters used in prior methods. Our code is publicly available: https://daveishan.github.io/codamal-webpage/.

Deep Metric Learning (DML) has long attracted the attention of the machine learning community as a key objective. Existing solutions concentrate on fine-tuning the pre-trained models on conventional image datasets. As a result of the success of recent pre-trained models derived from larger-scale datasets, it is challenging to adapt the model to the DML tasks in the local data domain while retaining the previously gained knowledge. In this paper, we investigate parameter-efficient methods for fine-tuning the pre-trained model for DML tasks. In particular, we propose a novel and effective framework based on learning Visual Prompts (VPT) in the pre-trained Vision Transformers (ViT). Based on the conventional proxy-based DML paradigm, we augment the proxy by incorporating the semantic information from the input image and the ViT, in which we optimize the visual prompts for each class. We demonstrate that our new approximations with semantic information are superior to representative capabilities, thereby improving metric learning performance. We conduct extensive experiments to demonstrate that our proposed framework is superior and efficient by evaluating popular DML benchmarks. In particular, we demonstrate that our fine-tuning method achieves comparable or even better performance than recent state-of-the-art full fine-tuning works of DML while tuning only a small percentage of total parameters.

3D detection is a critical task that enables machines to identify and locate objects in three-dimensional space. It has a broad range of applications in several fields, including autonomous driving, robotics and augmented reality. Monocular 3D detection is attractive as it requires only a single camera, however, it lacks the accuracy and robustness required for real world applications. High resolution LiDAR on the other hand, can be expensive and lead to interference problems in heavy traffic given their active transmissions. We propose a balanced approach that combines the advantages of monocular and point cloud-based 3D detection. Our method requires only a small number of 3D points, that can be obtained from a low-cost, low-resolution sensor. Specifically, we use only 512 points, which is just 1% of a full LiDAR frame in the KITTI dataset. Our method reconstructs a complete 3D point cloud from this limited 3D information combined with a single image. The reconstructed 3D point cloud and corresponding image can be used by any multi-modal off-the-shelf detector for 3D object detection. By using the proposed network architecture with an off-the-shelf multi-modal 3D detector, the accuracy of 3D detection improves by 20% compared to the state-of-theart monocular detection methods and 6% to 9% compare to the baseline multi-modal methods on KITTI and JackRabbot datasets.

Is the Text to Motion model robust? Recent advancements in Text to Motion models primarily stem from more accurate predictions of specific actions. However, the text modality typically relies solely on pre-trained Contrastive Language-Image Pretraining (CLIP) models. Our research has uncovered a significant issue with the text-tomotion model: its predictions often exhibit inconsistent outputs, resulting in vastly different or even incorrect poses when presented with semantically similar or identical text inputs. In this paper, we undertake an analysis to elucidate the underlying causes of this instability, establishing a clear link between the unpredictability of model outputs and the erratic attention patterns of the text encoder module. Consequently, we introduce a formal framework aimed at addressing this issue, which we term the Stable Text-to-Motion Framework (SATO). SATO consists of three modules, each dedicated to stable attention, stable prediction, and maintaining a balance between accuracy and robustness trade-off. We present a methodology for constructing an SATO that satisfies the stability of attention and prediction. To verify the stability of the model, we introduced a new textual synonym perturbation dataset based on HumanML3D and KIT-ML. Results show that SATO is significantly more stable against synonyms and other slight perturbations while keeping its high accuracy performance. Codes and models are released at

Grasp generation aims to create complex hand-object interactions with a specified object. While traditional approaches for hand generation have primarily focused on visibility and diversity under scene constraints, they tend to overlook the fine-grained hand-object interactions such as contacts, resulting in inaccurate and undesired grasps. To address these challenges, we propose a controllable grasp generation task and introduce ClickDiff, a controllable conditional generation model that leverages a fine-grained Semantic Contact Map (SCM). Particularly when synthesizing interactive grasps, the method enables the precise control of grasp synthesis through either user-specified or algorithmically predicted Semantic Contact Map. Specifically, to optimally utilize contact supervision constraints and to accurately model the complex physical structure of hands, we propose a Dual Generation Framework. Within this framework, the Semantic Conditional Module generates reasonable contact maps based on fine-grained contact information, while the Contact Conditional Module utilizes contact maps alongside object point clouds to generate realistic grasps. We evaluate the evaluation criteria applicable to controllable grasp generation. Both unimanual and bimanual generation experiments on GRAB and ARCTIC datasets verify the validity of our proposed method, demonstrating the efficacy and robustness of ClickDiff, even with previously unseen objects. Our code is available at https://github.com/adventurer-w/ClickDiff.

Skeleton-based action recognition has garnered significant attention due to the utilization of concise and resilient skeletons. Nevertheless, the absence of detailed body information in skeletons restricts performance, while other multimodal methods require substantial inference resources and are inefficient when using multimodal data during both training and inference stages. To address this and fully harness the complementary multimodal features, we propose a novel multi-modality co-learning (MMCL) framework by leveraging the multimodal large language models (LLMs) as auxiliary networks for efficient skeleton-based action recognition, which engages in multi-modality co-learning during the training stage and keeps efficiency by employing only concise skeletons in inference. Our MMCL framework primarily consists of two modules. First, the Feature Alignment Module (FAM) extracts rich RGB features from video frames and aligns them with global skeleton features via contrastive learning. Second, the Feature Refinement Module (FRM) uses RGB images with temporal information and text instruction to generate instructive features based on the powerful generalization of multimodal LLMs. These instructive text features will further refine the classification scores and the refined scores will enhance the model's robustness and generalization in a manner similar to soft labels. Extensive experiments on NTU RGB+D, NTU RGB+D 120 and Northwestern-UCLA benchmarks consistently verify the effectiveness of our MMCL, which outperforms the existing skeleton-based action recognition methods. Meanwhile, experiments on UTD-MHAD and SYSU-Action datasets demonstrate the commendable generalization of our MMCL in zero-shot and domain-adaptive action recognition. Our code is publicly available at: https://github.com/liujf69/MMCL-Action.

Recent work in Video Frame Interpolation (VFI) tries to formulate VFI as a diffusion-based conditional image generation problem, synthesizing the intermediate frame given a random noise and neighboring frames. Due to the relatively high resolution of videos, Latent Diffusion Models (LDMs) are employed to run diffusion models in latent space efficiently. Such a formulation poses a crucial challenge: VFI expects that the output is deterministically equal to the ground truth intermediate frame, but LDMs randomly generate a diverse set of different images when the model runs multiple times. The diversity is due to the large cumulative variance (variance accumulated at each generation step) of generated latent representations in LDMs, making the sampling trajectory random. To address this problem, we propose our unique solution: Frame Interpolation with Consecutive Brownian Bridge Diffusion. Specifically, we propose consecutive Brownian Bridge diffusion that takes a deterministic initial value as input, resulting in a much smaller cumulative variance of generated latent representations. Our experiments suggest that our method can improve together with the improvement of the autoencoder and achieve state-of-the-art performance in VFI, leaving strong potential for further enhancement. Our code is available at https://github.com/ZonglinL/ConsecutiveBrownianBridge.

Recently, transformers have demonstrated great potential for modeling long-term dependencies from skeleton sequences and thereby gained ever-increasing attention in skeleton action recognition. However, the existing transformer-based approaches heavily rely on the naive attention mechanism for capturing the spatiotemporal features, which falls short in learning discriminative representations that exhibit similar motion patterns. To address this challenge, we introduce the Frequency-aware Mixed Transformer (FreqMixFormer), specifically designed for recognizing similar skeletal actions with subtle discriminative motions. First, we introduce a frequency-aware attention module to unweave skeleton frequency representations by embedding joint features into frequency attention maps, aiming to distinguish the discriminative movements based on their frequency coefficients. Subsequently, we develop a mixed transformer architecture to incorporate spatial features with frequency features to model the comprehensive frequency-spatial patterns. Additionally, a temporal transformer is proposed to extract the global correlations across frames. Extensive experiments show that FreqMiXFormer outperforms SOTA on 3 popular skeleton action recognition datasets, including NTU RGB+D, NTU RGB+D 120, and NW-UCLA datasets. Our project is publicly available at: https://github.com/wenhanwu95/FreqMixFormer.

Federated Learning (FL) has emerged as a promising paradigm for collaborative machine learning, while preserving user data privacy. Despite its potential, standard FL algorithms lack support for diverse heterogeneous device prototypes, which vary significantly in model and dataset sizes---from small IoT devices to large workstations. This limitation is only partially addressed by existing knowledge distillation (KD) techniques, which often fail to transfer knowledge effectively across a broad spectrum of device prototypes with varied capabilities. This failure primarily stems from two issues: the dilution of informative logits from more capable devices by those from less capable ones, and the use of a single integrated logits as the distillation target across all devices, which neglects their individual learning capacities and and the unique contributions of each device. To address these challenges, we introduce TAKFL, a novel KD-based framework that treats the knowledge transfer from each device prototype's ensemble as a separate task, independently distilling each to preserve its unique contributions and avoid dilution. TAKFL also incorporates a KD-based self-regularization technique to mitigate the issues related to the noisy and unsupervised ensemble distillation process. To integrate the separately distilled knowledge, we introduce an adaptive task arithmetic knowledge integration process, allowing each student model to customize the knowledge integration for optimal performance. Additionally, we present theoretical results demonstrating the effectiveness of task arithmetic in transferring knowledge across heterogeneous device prototypes with varying capacities. Comprehensive evaluations of our method across both computer vision (CV) and natural language processing (NLP) tasks demonstrate that TAKFL achieves state-of-the-art results in a variety of datasets and settings, significantly outperforming existing KD-based methods. Our code is released at https://github.com/MMorafah/TAKFL and the project website is available at https://mmorafah.github.io/takflpage .

Federated learning (FL) allows collaborative machine learning training without sharing private data. While most FL methods assume identical data domains across clients, real-world scenarios often involve heterogeneous data domains. Federated Prototype Learning (FedPL) addresses this issue, using mean feature vectors as prototypes to enhance model generalization. However, existing FedPL methods create the same number of prototypes for each client, leading to cross-domain performance gaps and disparities for clients with varied data distributions. To mitigate cross-domain feature representation variance, we introduce FedPLVM, which establishes variance-aware dual-level prototypes clustering and employs a novel $\alpha$-sparsity prototype loss. The dual-level prototypes clustering strategy creates local clustered prototypes based on private data features, then performs global prototypes clustering to reduce communication complexity and preserve local data privacy. The $\alpha$-sparsity prototype loss aligns samples from underrepresented domains, enhancing intra-class similarity and reducing inter-class similarity. Evaluations on Digit-5, Office-10, and DomainNet datasets demonstrate our method's superiority over existing approaches.

Video Anomaly Detection (VAD) finds widespread applications in security surveillance, traffic monitoring, industrial monitoring, and healthcare. Despite extensive research efforts, there remains a lack of concise reviews that provide insightful guidance for researchers. Such reviews would serve as quick references to grasp current challenges, research trends, and future directions. In this paper, we present such a review, examining models and datasets from various perspectives. We emphasize the critical relationship between model and dataset, where the quality and diversity of datasets profoundly influence model performance, and dataset development adapts to the evolving needs of emerging approaches. Our review identifies practical issues, including the absence of comprehensive datasets with diverse scenarios. To address this, we introduce a new dataset, Multi-Scenario Anomaly Detection (MSAD), comprising 14 distinct scenarios captured from various camera views. Our dataset has diverse motion patterns and challenging variations, such as different lighting and weather conditions, providing a robust foundation for training superior models. We conduct an in-depth analysis of recent representative models using MSAD and highlight its potential in addressing the challenges of detecting anomalies across diverse and evolving surveillance scenarios.

Recent works have shown that the computational efficiency of 3D medical image (e.g. CT and MRI) segmentation can be impressively improved by dynamic inference based on slice-wise complexity. As a pioneering work, a dynamic architecture network for medical volumetric segmentation (i.e. Med-DANet [44]) has achieved a favorable accuracy and efficiency trade-off by dynamically selecting a suitable 2D candidate model from the pre-defined model bank for different slices. However, the issues of incomplete data analysis, high training costs, and the two-stage pipeline in Med-DANet require further improvement. To this end, this paper further explores a unified formulation of the dynamic inference framework from the perspective of both the data itself and the model structure. For each slice of the input volume, our proposed method dynamically selects an important foreground region for segmentation based on the policy generated by our Decision Network and Crop Position Network. Besides, we propose to insert a stage-wise quantization selector to the employed segmentation model (e.g. U-Net) for dynamic architecture adapting. Extensive experiments on BraTS 2019 and 2020 show that our method achieves comparable or better performance than previous state-of-the-art methods with much less model complexity. Compared with previous methods Med-DANet and TransBTS with dynamic and static architecture respectively, our framework improves the model efficiency by up to nearly 4.1 and 17.3 times with comparable segmentation results on BraTS 2019. Code will be available at https://github.com/Rubics-Xuan/Med-DANet.

The research community has witnessed the powerful potential of self-supervised Masked Image Modeling (MIM), which enables the models capable of learning visual representation from unlabeled data. In this paper, to incorporate both the crucial global structural information and local details for dense prediction tasks, we alter the perspective to the frequency domain and present a new MIM-based framework named FreMIM for self-supervised pre-training to better accomplish medical image segmentation tasks. Based on the observations that the detailed structural information mainly lies in the high-frequency components and the high-level semantics are abundant in the low-frequency counterparts, we further incorporate multi-stage supervision to guide the representation learning during the pre-training phase. Extensive experiments on three benchmark datasets show the superior advantage of our FreMIM over previous state-of-the-art MIM methods. Compared with various baselines trained from scratch, our FreMIM could consistently bring considerable improvements to model performance. The code will be publicly available at https://github.com/jingw193/FreMIM.

Cross-View Geo-Localization (CVGL) estimates the location of a ground image by matching it to a geo-tagged aerial image in a database. Recent works achieve outstanding progress on CVGL benchmarks. However, existing methods still suffer from poor performance in cross-area evaluation, in which the training and testing data are captured from completely distinct areas. We attribute this deficiency to the lack of ability to extract the geometric layout of visual features and models' overfitting to low-level details. Our preliminary work (Zhang et al. 2022) introduced a Geometric Layout Extractor (GLE) to capture the geometric layout from input features. However, the previous GLE does not fully exploit information in the input feature. In this work, we propose GeoDTR+ with an enhanced GLE module that better models the correlations among visual features. To fully explore the LS techniques from our preliminary work, we further propose Contrastive Hard Samples Generation (CHSG) to facilitate model training. Extensive experiments show that GeoDTR+ achieves state-of-the-art (SOTA) results in cross-area evaluation on CVUSA (Workman et al. 2015), CVACT (Liu and Li, 2019), and VIGOR (Zhu et al. 2021) by a large margin (16.44%, 22.71%, and 13.66% without polar transformation) while keeping the same-area performance comparable to existing SOTA. Moreover, we provide detailed analyses of GeoDTR+.

Source code authorship attribution is an important problem in practical applications such as plagiarism detection, software forensics, and copyright disputes. Recent studies show that existing methods for source code authorship attribution can be significantly affected by time evolution, leading to a decrease in attribution accuracy year by year. To alleviate the problem of Deep Learning (DL)-based source code authorship attribution degrading in accuracy due to time evolution, we propose a new framework called Time D omain A daptation (TimeDA) by adding new feature extractors to the original DL-based code attribution framework that enhances the learning ability of the original model on source domain features without requiring new or more source data. Moreover, we employ a centroid-based pseudo-labeling strategy using neighborhood clustering entropy for adaptive learning to improve the robustness of DL-based code authorship attribution. Experimental results show that TimeDA can significantly enhance the robustness of DL-based source code authorship attribution to time evolution, with an average improvement of 8.7% on the Java dataset and 5.2% on the C++ dataset. In addition, our TimeDA benefits from employing the centroid-based pseudo-labeling strategy, which significantly reduced the model training time by 87.3% compared to traditional unsupervised domain adaptive methods.

Most skeleton-based action recognition methods assume that the same type of action samples in the training set and the test set share similar motion patterns. However, action samples in real scenarios usually contain novel motion patterns which are not involved in the training set. As it is laborious to collect sufficient training samples to enumerate various types of novel motion patterns, this paper presents a practical skeleton-based action recognition task where the training set contains common motion patterns of action samples and the test set contains action samples that suffer from novel motion patterns. For this task, we present a Mask Graph Convolutional Network (Mask-GCN) to focus on learning action-specific skeleton joints that mainly convey action information meanwhile masking action-agnostic skeleton joints that convey rare action information and suffer more from novel motion patterns. Specifically, we design a policy network to learn layer-wise body masks to construct masked adjacency matrices, which guide a GCN-based backbone to learn stable yet informative action features from dynamic graph structure. Extensive experiments on our newly collected dataset verify that Mask-GCN outperforms most GCN-based methods when testing with various novel motion patterns.

Clustered federated learning (FL) has been shown to produce promising results by grouping clients into clusters. This is especially effective in scenarios where separate groups of clients have significant differences in the distributions of their local data. Existing clustered FL algorithms are essentially trying to group together clients with similar distributions so that clients in the same cluster can leverage each other's data to better perform federated learning. However, prior clustered FL algorithms attempt to learn these distribution similarities indirectly during training, which can be quite time consuming as many rounds of federated learning may be required until the formation of clusters is stabilized. In this paper, we propose a new approach to federated learning that directly aims to efficiently identify distribution similarities among clients by analyzing the principal angles between the client data subspaces. Each client applies a truncated singular value decomposition (SVD) step on its local data in a single-shot manner to derive a small set of principal vectors, which provides a signature that succinctly captures the main characteristics of the underlying distribution. This small set of principal vectors is provided to the server so that the server can directly identify distribution similarities among the clients to form clusters. This is achieved by comparing the similarities of the principal angles between the client data subspaces spanned by those principal vectors. The approach provides a simple, yet effective clustered FL framework that addresses a broad range of data heterogeneity issues beyond simpler forms of Non-IIDness like label skews. Our clustered FL approach also enables convergence guarantees for non-convex objectives.

Video captioning aims to generate natural language sentences that describe the given video accurately. Existing methods obtain favorable generation by exploring richer visual representations in encode phase or improving the decoding ability. However, the long-tailed problem hinders these attempts at low-frequency tokens, which rarely occur but carry critical semantics, playing a vital role in the detailed generation. In this paper, we introduce a novel Refined Semantic enhancement method towards Frequency Diffusion (RSFD), a captioning model that constantly perceives the linguistic representation of the infrequent tokens. Concretely, a Frequency-Aware Diffusion (FAD) module is proposed to comprehend the semantics of low-frequency tokens to break through generation limitations. In this way, the caption is refined by promoting the absorption of tokens with insufficient occurrence. Based on FAD, we design a Divergent Semantic Supervisor (DSS) module to compensate for the information loss of high-frequency tokens brought by the diffusion process, where the semantics of low-frequency tokens is further emphasized to alleviate the long-tailed problem. Extensive experiments indicate that RSFD outperforms the state-of-the-art methods on two benchmark datasets, i.e., MSR-VTT and MSVD, demonstrate that the enhancement of low-frequency tokens semantics can obtain a competitive generation effect. Code is available at https://github.com/lzp870/RSFD.

Semi-Supervised Learning can be more beneficial for the video domain compared to images because of its higher an-notation cost and dimensionality. Besides, any video understanding task requires reasoning over both spatial and temporal dimensions. In order to learn both the static and motion related features for the semi-supervised action recognition task, existing methods rely on hard in-put inductive biases like using two-modalities (RGB and Optical-flow) or two-stream of different playback rates. Instead of utilizing unlabeled videos through diverse in-put streams, we rely on self-supervised video represen-tations, particularly, we utilize temporally-invariant and temporally-distinctive representations. We observe that these representations complement each other depending on the nature of the action. Based on this observation, we propose a student-teacher semi-supervised learning frame-work, TimeBalance, where we distill the knowledge from a temporally-invariant and a temporally-distinctive teacher. Depending on the nature of the unlabeled video, we dy-namically combine the knowledge of these two teach-ers based on a novel temporal similarity-based reweighting scheme. Our method achieves state-of-the-art performance on three action recognition benchmarks: UCF101, HMDB51, and Kinetics400. Code: https://github.com/DAVEISHAN/TimeBalance.

Recently, transformer-based methods have gained significant success in sequential 2D-to-3D lifting human pose estimation. As a pioneering work, PoseFormer captures spatial relations of human joints in each video frame and human dynamics across frames with cascaded transformer layers and has achieved impressive performance. However, in real scenarios, the performance of PoseFormer and its follow-ups is limited by two factors: (a) The length of the input joint sequence; (b) The quality of 2D joint detection. Existing methods typically apply self-attention to all frames of the input sequence, causing a huge computational burden when the frame number is increased to obtain advanced estimation accuracy, and they are not robust to noise naturally brought by the limited capability of 2D joint detectors. In this paper, we propose PoseFormerV2, which exploits a compact representation of lengthy skeleton sequences in the frequency domain to efficiently scale up the receptive field and boost robustness to noisy 2D joint detection. With minimum modifications to PoseFormer, the proposed method effectively fuses features both in the time domain and frequency domain, enjoying a better speed-accuracy trade-off than its precursor. Extensive experiments on two benchmark datasets (i.e., Human3.6M and MPI-INF-3DHP) demonstrate that the proposed approach significantly outperforms the original PoseFormer and other transformer-based variants. Code is released at https://github.com/ QitaoZhao/PoseFormerV2.

Transformer architectures have achieved SOTA performance on the human mesh recovery (HMR) from monocular images. However, the performance gain has come at the cost of substantial memory and computational overhead. A lightweight and efficient model to reconstruct accurate human mesh is needed for real-world applications. In this paper, we propose a pure transformer architecture named POoling aTtention TransformER (POTTER) for the HMR task from single images. Observing that the conventional attention module is memory and computationally expensive, we propose an efficient pooling attention module, which significantly reduces the memory and computational cost without sacrificing performance. Furthermore, we design a new transformer architecture by integrating a High-Resolution (HR) stream for the HMR task. The high-resolution local and global features from the HR stream can be utilized for recovering more accurate human mesh. Our POTTER outperforms the SOTA method METRO by only requiring 7% of total parameters and 14% of the Multiply-Accumulate Operations on the Human3.6M (PA-MPJPE metric) and 3DPW (all three metrics) datasets. The project webpage is https://zczcwh.github.io/potter_page/.

Recently, vision transformers have shown great success in a set of human reconstruction tasks such as 2D/3D human pose estimation (2D/3D HPE) and human mesh reconstruction (HMR) tasks. In these tasks, feature map representations of the human structural information are often extracted first from the image by a CNN (such as HRNet), and then further processed by transformer to predict the heatmaps for HPE or HMR. However, existing transformer architectures are not able to process these feature map inputs directly, forcing an unnatural flattening of the location-sensitive human structural information. Furthermore, much of the performance benefit in recent HPE and HMR methods has come at the cost of ever-increasing computation and memory needs. Therefore, to simultaneously address these problems, we propose FeatER, a novel transformer design that preserves the inherent structure of feature map representations when modeling attention while reducing memory and computational costs. Taking advantage of FeatER, we build an efficient network for a set of human reconstruction tasks including 2D HPE, 3D HPE, and HMR. A feature map reconstruction module is applied to improve the performance of the estimated human pose and mesh. Extensive experiments demonstrate the effectiveness of FeatER on various human pose and mesh datasets. For instance, FeatER outperforms the SOTA method Mesh- Graphormer by requiring 5% of Params and 16% of MACs on Human3.6M and 3DPW datasets. The project webpage is https://zczcwh.github.io/feater_page/.

We investigate the problem of automatically placing an object into a background image for image compositing. Given a background image and a segmented object, the goal is to train a model to predict plausible placements (location and scale) of the object for compositing. The quality of the composite image highly depends on the predicted location/scale. Existing works either generate candidate bounding boxes or apply sliding-window search using global representations from background and object images, which fail to model local information in background images. However, local clues in background images are important to determine the compatibility of placing the objects with certain locations/scales. In this paper, we propose to learn the correlation between object features and all local background features with a transformer module so that detailed information can be provided on all possible location/scale configurations. A sparse contrastive loss is further proposed to train our model with sparse supervision. Our new formulation generates a 3D heatmap indicating the plausibility of all location/scale combinations in one network forward pass, which is > 10 x faster than the previous sliding-window method. It also supports interactive search when users provide a pre-defined location or scale. The proposed method can be trained with explicit annotation or in a self-supervised manner using an off-the-shelf inpainting model, and it outperforms state-of-the-art methods significantly. User study shows that the trained model generalizes well to real-world images with diverse challenging scenes and object categories.

Visual Place Recognition (VPR) estimates the location of query images by matching them with images in a reference database. Conventional methods generally adopt aggregated CNN features for global retrieval and RANSAC-based geometric verification for reranking. However, RANSAC only employs geometric information but ignores other possible information that could be useful for reranking, e.g. local feature correlations, and attention values. In this paper, we propose a unified place recognition framework that handles both retrieval and reranking with a novel transformer model, named$R^{2}$Former. The proposed reranking module takes feature correlation, attention value, and$xy$coordinates into account, and learns to determine whether the image pair is from the same location. The whole pipeline is end-to-end trainable and the reranking module alone can also be adopted on other CNN or transformer backbones as a generic component. Remarkably,$R^{2}$Former significantly outperforms state-of-the-art methods on major VPR datasets with much less inference time and memory consumption. It also achieves the state-of-the-art on the hold-out MSLS challenge set and could serve as a simple yet strong solution for real-world large-scale applications. Experiments also show vision transformer tokens are comparable and sometimes better than CNN local features on local matching. The code is released at https://github.com/Jeff-Zilence/R2Former.

Most existing gait recognition methods are appearance-based, which rely on the silhouettes extracted from the video data of human walking activities. The less-investigated skeleton-based gait recognition methods directly learn the gait dynamics from 2D/3D human skeleton sequences, which are theoretically more robust solutions in the presence of appearance changes caused by clothes, hairstyles, and carrying objects. However, the performance of skeleton-based solutions is still largely behind the appearance-based ones. This paper aims to close such performance gap by proposing a novel network model, GaitMixer, to learn more discriminative gait representation from skeleton sequence data. In particular, GaitMixer follows a heterogeneous multi-axial mixer architecture, which exploits the spatial self-attention mixer followed by the temporal large-kernel convolution mixer to learn rich multi-frequency signals in the gait feature maps. Experiments on the widely used gait database, CASIA-B, demonstrate that GaitMixer outperforms the previous SOTA skeleton-based methods by a large margin while achieving a competitive performance compared with the representative appearance-based solutions. Code will be available at https://github.com/exitudio/gaitmixer

With its significant performance improvements, the deep learning paradigm has become a standard tool for modern image denoisers. While promising performance has been shown on seen noise distributions, existing approaches often suffer from generalisation to unseen noise types or general and real noise. It is understandable as the model is designed to learn paired mapping (e.g. from a noisy image to its clean version). In this paper, we instead propose to learn to disentangle the noisy image, under the intuitive assumption that different corrupted versions of the same clean image share a common latent space. A self-supervised learning framework is proposed to achieve the goal, without looking at the latent clean image. By taking two different corrupted versions of the same image as input, the proposed Multi-view Self-supervised Disentanglement (MeD) approach learns to disentangle the latent clean features from the corruptions and recover the clean image consequently. Extensive experimental analysis on both synthetic and real noise shows the superiority of the proposed method over prior self-supervised approaches, especially on unseen novel noise types. On real noise, the proposed method even outperforms its supervised counterparts by over 3 dB.

The goal of building a benchmark (suite of datasets) is to provide a unified protocol for fair evaluation and thus facilitate the evolution of a specific area. Nonetheless, we point out that existing protocols of action recognition could yield partial evaluations due to several limitations. To comprehensively probe the effectiveness of spatiotemporal representation learning, we introduce BEAR, a new BEnchmark on video Action Recognition. BEAR is a collection of 18 video datasets grouped into 5 categories (anomaly, gesture, daily, sports, and instructional), which covers a diverse set of real-world applications. With BEAR, we thoroughly evaluate 6 common spatiotemporal models pre-trained by both supervised and self-supervised learning. We also report transfer performance via standard finetuning, few-shot finetuning, and unsupervised domain adaptation. Our observation suggests that the current state-of-the-art cannot solidly guarantee high performance on datasets close to real-world applications, and we hope BEAR can serve as a fair and challenging evaluation benchmark to gain insights on building next-generation spatiotemporal learners. Our dataset, code, and models are released at: https://github.com/AndongDeng/BEAR

Pseudo-labeling is significant for semi-supervised instance segmentation, which generates instance masks and classes from unannotated images for subsequent training. However, in existing pipelines, pseudo-labels that contain valuable information may be directly filtered out due to mismatches in class and mask quality. To address this issue, we propose a novel framework, called pseudo-label aligning instance segmentation (PAIS), in this paper. In PAIS, we devise a dynamic aligning loss (DALoss) that adjusts the weights of semi-supervised loss terms with varying class and mask score pairs. Through extensive experiments conducted on the COCO and Cityscapes datasets, we demonstrate that PAIS is a promising framework for semi-supervised instance segmentation, particularly in cases where labeled data is severely limited. Notably, with just 1% labeled data, PAIS achieves 21.2 mAP (based on MaskRCNN) and 19.9 mAP (based on K-Net) on the COCO dataset, outperforming the current state-of-the-art model, i.e., NoisyBoundary with 7.7 mAP, by a margin of over 12 points. Code is available at: https://github.com/hujiecpp/PAIS.

The current interacting hand (IH) datasets are relatively simplistic in terms of background and texture, with hand joints being annotated by a machine annotator, which may result in inaccuracies, and the diversity of pose distribution is limited. However, the variability of background, pose distribution, and texture can greatly influence the generalization ability. Therefore, we present a large-scale synthetic dataset –RenderIH– for interacting hands with accurate and diverse pose annotations. The dataset contains 1M photo-realistic images with varied backgrounds, perspectives, and hand textures. To generate natural and diverse interacting poses, we propose a new pose optimization algorithm. Additionally, for better pose estimation accuracy, we introduce a transformer-based pose estimation network, TransHand, to leverage the correlation between interacting hands and verify the effectiveness of RenderIH in improving results. Our dataset is model-agnostic and can improve more accuracy of any hand pose estimation method in comparison to other real or synthetic datasets. Experiments have shown that pretraining on our synthetic data can significantly decrease the error from 6.76mm to 5.79mm, and our Transhand surpasses contemporary methods. Our dataset and code are available at https://github.com/adwardlee/RenderIH.

The paradigm of large-scale pre-training followed by downstream fine-tuning has been widely employed in various object detection algorithms. In this paper, we reveal discrepancies in data, model, and task between the pre-training and fine-tuning procedure in existing practices, which implicitly limit the detector’s performance, generalization ability, and convergence speed. To this end, we propose AlignDet, a unified pre-training framework that can be adapted to various existing detectors to alleviate the discrepancies. AlignDet decouples the pre-training process into two stages, i.e., image-domain and box-domain pre-training. The image-domain pre-training optimizes the detection backbone to capture holistic visual abstraction, and box-domain pre-training learns instance-level semantics and task-aware concepts to initialize the parts out of the backbone. By incorporating the self-supervised pretrained backbones, we can pre-train all modules for various detectors in an unsupervised paradigm. As depicted in Figure 1, extensive experiments demonstrate that AlignDet can achieve significant improvements across diverse protocols, such as ${\color{Green}\text{detection algorithms}}, {\color{Blue}\text{model backbones}}, {\color{Red}\text{data settings}}$, and ${\color{SkyBlue}\text{training schedules}}$. For example, AlignDet improves FCOS by 5.3 mAP, RetinaNet by 2.1 mAP, Faster R-CNN by 3.3 mAP, and DETR by 2.3 mAP under fewer epochs.

Personalized federated learning has received an upsurge of attention due to the mediocre performance of conventional federated learning (FL) over heterogeneous data. Unlike conventional FL which trains a single global consensus model, personalized FL allows different models for different clients. However, existing personalized FL algorithms only implicitly transfer the collaborative knowledge across the federation by embedding the knowledge into the aggregated model or regularization. We observed that this implicit knowledge transfer fails to maximize the potential of each client’s empirical risk toward other clients. Based on our observation, in this work, we propose Personalized Global Federated Learning (PGFed), a novel personalized FL framework that enables each client to personalize its own global objective by explicitly and adaptively aggregating the empirical risks of itself and other clients. To avoid massive (O(N2)) communication overhead and potential privacy leakage while achieving this, each client’s risk is estimated through a first-order approximation for other clients’ adaptive risk aggregation. On top of PGFed, we develop a momentum upgrade, dubbed PGFedMo, to more efficiently utilize clients’ empirical risks. Our extensive experiments on four datasets under different federated settings show consistent improvements of PGFed over previous state-of-the-art methods. The code is publicly available at https://github.com/ljaiverson/pgfed.

Geospatial technologies are becoming increasingly essential in our world for a wide range of applications, including agriculture, urban planning, and disaster response. To help improve the applicability and performance of deep learning models on these geospatial tasks, various works have begun investigating foundation models for this domain. Researchers have explored two prominent approaches for introducing such models in geospatial applications, but both have drawbacks in terms of limited performance benefit or prohibitive training cost. Therefore, in this work, we propose a novel paradigm for building highly effective geospatial foundation models with minimal resource cost and carbon impact. We first construct a compact yet diverse dataset from multiple sources to promote feature diversity, which we term GeoPile. Then, we investigate the potential of continual pretraining from large-scale ImageNet-22k models and propose a multi-objective continual pretraining paradigm, which leverages the strong representations of ImageNet while simultaneously providing the freedom to learn valuable in-domain features. Our approach outperforms previous state-of-the-art geospatial pretraining methods in an extensive evaluation on seven downstream datasets covering various tasks such as change detection, classification, multi-label classification, semantic segmentation, and super-resolution. Code is available at https://github.com/mmendiet/GFM.

Human Pose Estimation (HPE) is widely used in various fields, including motion analysis, healthcare, and virtual reality. However, the great expenses of labeled real-world datasets present a significant challenge for HPE. To overcome this, one approach is to train HPE models on synthetic datasets and then perform domain adaptation (DA) on real-world data. Unfortunately, existing DA methods for HPE neglect data privacy and security by using both source and target data in the adaptation process.To this end, we propose a new task, named source-free domain adaptive HPE, which aims to address the challenges of cross-domain learning of HPE without access to source data during the adaptation process. We further propose a novel framework that consists of three models: source model, intermediate model, and target model, which explores the task from both source-protect and target-relevant perspectives. The source-protect module preserves source information more effectively while resisting noise, and the target-relevant module reduces the sparsity of spatial representations by building a novel spatial probability space, and pose-specific contrastive learning and information maximization are proposed on the basis of this space. Comprehensive experiments on several domain adaptive HPE benchmarks show that the proposed method outperforms existing approaches by a considerable margin. The codes are available at https://github.com/davidpengucf/SFDAHPE.

Personalized Federated Learning (PFL) represents a promising solution for decentralized learning in heterogeneous data environments. Partial model personalization has been proposed to improve the efficiency of PFL by selectively updating local model parameters instead of aggregating all of them. However, previous work on partial model personalization has mainly focused on Convolutional Neural Networks (CNNs), leaving a gap in understanding how it can be applied to other popular models such as Vision Transformers (ViTs). In this work, we investigate where and how to partially personalize a ViT model. Specifically, we empirically evaluate the sensitivity to data distribution of each type of layer. Based on the insights that the self-attention layer and the classification head are the most sensitive parts of a ViT, we propose a novel approach called FedPerfix, which leverages plugins to transfer information from the aggregated model to the local client as a personalization. Finally, we evaluate the proposed approach on CIFAR-100, OrganAMNIST, and Office-Home datasets and demonstrate its effectiveness in improving the model’s performance compared to several advanced PFL methods. Code is available at https://github.com/imguangyu/FedPerfix

An oft-cited open problem of federated learning is the existence of data heterogeneity among clients. One pathway to understanding the drastic accuracy drop in federated learning is by scrutinizing the behavior of the clients’ deep models on data with different levels of "difficulty", which has been left unaddressed. In this paper, we investigate a different and rarely studied dimension of FL: ordered learning. Specifically, we aim to investigate how ordered learning principles can contribute to alleviating the heterogeneity effects in FL. We present theoretical analysis and conduct extensive empirical studies on the efficacy of orderings spanning three kinds of learning: curriculum, anti-curriculum, and random curriculum. We find that curriculum learning largely alleviates non-IIDness. Interestingly, the more disparate the data distributions across clients the more they benefit from ordered learning. We provide analysis explaining this phenomenon, specifically indicating how curriculum training appears to make the objective landscape progressively less convex, suggesting fast converging iterations at the beginning of the training procedure. We derive quantitative results of convergence for both convex and nonconvex objectives by modeling the curriculum training on federated devices as local SGD with locally biased stochastic gradients. Also, inspired by ordered learning, we propose a novel client selection technique that benefits from the real-world disparity in the clients. Our proposed approach to client selection has a synergic effect when applied together with ordered learning in FL.

In recent years, remarkable results have been achieved in self-supervised action recognition using skeleton sequences with contrastive learning. It has been observed that the semantic distinction of human action features is often represented by local body parts, such as legs or hands, which are advantageous for skeleton-based action recognition. This paper proposes an attention-based contrastive learning framework for skeleton representation learning, called SkeAttnCLR, which integrates local similarity and global features for skeleton-based action representations. To achieve this, a multi-head attention mask module is employed to learn the soft attention mask features from the skeletons, suppressing non-salient local features while accentuating local salient features, thereby bringing similar local features closer in the feature space. Additionally, ample contrastive pairs are generated by expanding contrastive pairs based on salient and non-salient features with global features, which guide the network to learn the semantic representations of the entire skeleton. Therefore, with the attention mask mechanism, SkeAttnCLR learns local features under different data augmentation views. The experiment results demonstrate that the inclusion of local feature similarity significantly enhances skeleton-based action representation. Our proposed SkeAttnCLR outperforms state-of-the-art methods on NTURGB+D, NTU120-RGB+D, and PKU-MMD datasets. The code and settings are available at this repository: https://github.com/GitHubOfHyl97/SkeAttnCLR.

Human-centric visual understanding is an important desideratum for effective human-robot interaction. In order to navigate crowded public places, social robots must be able to interpret the activity of the surrounding humans. This paper addresses one key aspect of human-centric visual understanding, multi-person pose estimation. Achieving good performance on multi-person pose estimation in crowded scenes is difficult due to the challenges of occluded joints and instance separation. In order to tackle these challenges and overcome the limitations of image features in representing invisible body parts, we propose a novel prompt-based pose inference strategy called LAMP (Language Assisted Multi-person Pose estimation). By utilizing the text representations generated by a well-trained language model (CLIP), LAMP can facilitate the understanding of poses on the instance and joint levels, and learn more robust visual representations that are less susceptible to occlusion. This paper demonstrates that language-supervised training boosts the performance of single-stage multi-person pose estimation, and both instance-level and joint-level prompts are valuable for training. The code is available at https://github.com/shengnanh20/LAMP.

Human-robot interaction (HRI) is a rapidly growing field that encompasses social and industrial applications. Machine learning plays a vital role in industrial HRI by enhancing the adaptability and autonomy of robots in complex environments. However, data privacy is a crucial concern in the interaction between humans and robots, as companies need to protect sensitive data while machine learning algorithms require access to large datasets. Federated Learning (FL) offers a solution by enabling the distributed training of models without sharing raw data. Despite extensive research on Federated learning (FL) for tasks such as natural language processing (NLP) and image classification, the question of how to use FL for HRI remains an open research problem. The traditional FL approach involves transmitting large neural network parameter matrices between the server and clients, which can lead to high communication costs and often becomes a bottleneck in FL. This paper proposes a communication-efficient FL framework for human-robot interaction (CEFHRI) to address the challenges of data heterogeneity and communication costs. The framework leverages pre-trained models and introduces a trainable spatiotemporal adapter for video understanding tasks in HRI. Experimental results on three human-robot interaction benchmark datasets: HRI30, InHARD, and COIN demonstrate the superiority of CEFHRI over full fine-tuning in terms of communication costs. The proposed methodology provides a secure and efficient approach to HRI federated learning, particularly in industrial environments with data privacy concerns and limited communication bandwidth. Our code is available at https://github.com/umarkhalidAI/CEFHRI-Efficient-Federated-Learning.

Deep learning has been widely used in source code classification tasks, such as code classification according to their functionalities, code authorship attribution, and vulnerability detection. Unfortunately, the black-box nature of deep learning makes it hard to interpret and understand why a classifier (i.e., classification model) makes a particular prediction on a given example. This lack of interpretability (or explainability) might have hindered their adoption by practitioners because it is not clear when they should or should not trust a classifier's prediction. The lack of interpretability has motivated a number of studies in recent years. However, existing methods are neither robust nor able to cope with out-of-distribution examples. In this paper, we propose a novel method to produce Robust interpreters for a given deep learning-based code classifier; the method is dubbed Robin. The key idea behind Robin is a novel hybrid structure combining an interpreter and two approximators, while leveraging the ideas of adversarial training and data augmentation. Experimental results show that on average the interpreter produced by Robin achieves a 6.11% higher fidelity (evaluated on the classifier), 67.22% higher fidelity (evaluated on the approximator), and 15.87x higher robustness than that of the three existing interpreters we evaluated. Moreover, the interpreter is 47.31% less affected by out-of-distribution examples than that of LEMNA.

Masked Autoencoders (MAE) have demonstrated promising performance in self-supervised learning for both 2D and 3D computer vision. Nevertheless, existing MAE-based methods still have certain drawbacks. Firstly, the functional decoupling between the encoder and decoder is incomplete, which limits the encoder's representation learning ability. Secondly, downstream tasks solely utilize the encoder, failing to fully leverage the knowledge acquired through the encoder-decoder architecture in the pre-text task. In this paper, we propose Point Regress AutoEncoder (Point-RAE), a new scheme for regressive autoencoders for point cloud self-supervised learning. The proposed method decouples functions between the decoder and the encoder by introducing a mask regressor, which predicts the masked patch representation from the visible patch representation encoded by the encoder and the decoder reconstructs the target from the predicted masked patch representation. By doing so, we minimize the impact of decoder updates on the representation space of the encoder. Moreover, we introduce an alignment constraint to ensure that the representations for masked patches, predicted from the encoded representations of visible patches, are aligned with the masked patch presentations computed from the encoder. To make full use of the knowledge learned in the pre-training stage, we design a new finetune mode for the proposed Point-RAE. Extensive experiments demonstrate that our approach is efficient during pre-training and generalizes well on various downstream tasks. Specifically, our pre-trained models achieve a high accuracy of 90.28% on the ScanObjectNN hardest split and 94.1% accuracy on ModelNet40, surpassing all the other self-supervised learning methods. Our code and pretrained model are public available at: https://github.com/liuyyy111/Point-RAE.

Existing Graph Convolutional Networks to achieve human motion prediction largely adopt a one-step scheme, which output the prediction straight from history input, failing to exploit human motion patterns. We observe that human motions have transitional patterns and can be split into snippets representative of each transition. Each snippet can be reconstructed from its starting and ending poses referred to as the transitional poses. We propose a snippet-to-motion multi-stage framework that breaks motion prediction into sub-tasks easier to accomplish. Each sub-task integrates three modules: transitional pose prediction, snippet reconstruction, and snippet-to-motion prediction. Specifically, we propose to first predict only the transitional poses. Then we use them to reconstruct the corresponding snippets, obtaining a close approximation to the true motion sequence. Finally we refine them to produce the final prediction output. To implement the network, we propose a novel unified graph modeling, which allows for direct and effective feature propagation compared to existing approaches which rely on separate space-time modeling. Extensive experiments on Human 3.6M, CMU Mocap and 3DPW datasets verify the effectiveness of our method which achieves state-of-the-art performance.

In recent years, Graph Convolutional Networks (GCNs) have been widely used in human motion prediction, but their performance remains unsatisfactory. Recently, MLP-Mixer, initially developed for vision tasks, has been leveraged into human motion prediction as a promising alternative to GCNs, which achieves both better performance and better efficiency than GCNs.

Recent neural architecture search (NAS) works proposed training-free metrics to rank networks which largely reduced the search cost in NAS. In this paper, we revisit these training-free metrics and find that: (1) the number of parameters (#Param), which is the most straightforward training-free metric, is overlooked in previous works but is surprisingly effective, (2) recent training-free metrics largely rely on the #Param information to rank networks. Our experiments show that the performance of recent training-free metrics drops dramatically when the #Param information is not available. Motivated by these observations, we argue that metrics less correlated with the #Param are desired to provide additional information for NAS. We propose a light-weight training-based metric which has a weak correlation with the #Param while achieving better performance than training-free metrics at a lower search cost. Specifically, on DARTS search space, our method completes searching directly on ImageNet in only 2.6 GPU hours and achieves a top-1/top-5 error rate of 24.1%/7.1%, which is competitive among state-of-the-art NAS methods.

Most existing deep neural networks are static, which means they can only perform inference at a fixed complexity. But the resource budget can vary substantially across different devices. Even on a single device, the affordable budget can change with different scenarios, and repeatedly training networks for each required budget would be incredibly expensive. Therefore, in this work, we propose a general method called MutualNet to train a single network that can run at a diverse set of resource constraints. Our method trains a cohort of model configurations with various network widths and input resolutions. This mutual learning scheme not only allows the model to run at different width-resolution configurations but also transfers the unique knowledge among these configurations, helping the model to learn stronger representations overall. MutualNet is a general training methodology that can be applied to various network structures (e.g., 2D networks: MobileNets, ResNet, 3D networks: SlowFast, X3D) and various tasks (e.g., image classification, object detection, segmentation, and action recognition), and is demonstrated to achieve consistent improvements on a variety of datasets. Since we only train the model once, it also greatly reduces the training cost compared to independently training several models. Surprisingly, MutualNet can also be used to significantly boost the performance of a single network, if dynamic resource constraints are not a concern. In summary, MutualNet is a unified method for both static and adaptive, 2D and 3D networks. Code and pre-trained models are available at https://github.com/taoyang1122/MutualNet.

Deep learning models have been able to generate rain-free images effectively, but the extension of these methods to complex rain conditions where rain streaks show various blurring degrees, shapes, and densities has remained an open problem. Among the major challenges are the capacity to encode the rain streaks and the sheer difficulty of learning multi-scale context features that preserve both global color coherence and exactness of detail. To address the first problem, we design a non-local fusion module (NFM) and an attention fusion module (AFM), and construct the multi-level pyramids' architecture to explore the local and global correlations of rain information from the rain image pyramid. More specifically, we apply the non-local operation to fully exploit the self-similarity of rain streaks and perform the fusion of multi-scale features along the image pyramid. To address the latter challenge, we additionally design a residual learning branch that is capable of adaptively bridging the gaps (e.g., texture and color information) between the predicted rain-free image and the clean background via a hybrid embedding representation. Extensive results have demonstrated that our proposed method is able to generate much better rain-free images on several benchmark datasets than the state-of-the-art algorithms. Moreover, we conduct the joint evaluation experiments with respect to deraining performance and the detection/segmentation accuracy to further verify the effectiveness of our deraining method for downstream vision tasks/applications. The source code is available at https://github.com/kuihua/MSHFN.

Low-light image enhancement aims to improve an image's visibility while keeping its visual naturalness. Different from existing methods, which tend to accomplish the relighting task directly, we investigate the intrinsic degradation and relight the low-light image while refining the details and color in two steps. Inspired by the color image formulation (diffuse illumination color plus environment illumination color), we first estimate the degradation from low-light inputs to simulate the distortion of environment illumination color, and then refine the content to recover the loss of diffuse illumination color. To this end, we propose a novel Degradation-to-Refinement Generation Network (DRGN). Its distinctive features can be summarized as 1) A novel two-step generation network for degradation learning and content refinement. It is not only superior to one-step methods, but also capable of synthesizing sufficient paired samples to benefit the model training; 2) A multi-resolution fusion network to represent the target information (degradation or contents) in a multi-scale cooperative manner, which is more effective to address the complex unmixing problems. Extensive experiments on both the enhancement task and the joint detection task have verified the effectiveness and efficiency of our proposed method, surpassing the SOTA by 1.59dB on average and 3.18\% in mAP on the ExDark dataset. The code will be available soon.

Visual private information leakage is an emerging key is-sue for the fast growing applications of video understanding like activity recognition. Existing approaches for mitigating privacy leakage in action recognition require privacy labels along with the action labels from the video dataset. However, annotating frames of video dataset for privacy la-bels is not feasible. Recent developments of self-supervised learning (SSL) have unleashed the untapped potential of the unlabeled data. For the first time, we present a novel training framework which removes privacy information from in-put video in a self-supervised manner without requiring pri-vacy labels. Our training framework consists of three main components: anonymization function, self-supervised pri-vacy removal branch, and action recognition branch. We train our framework using a minimax optimization strategy to minimize the action recognition cost function and max-imize the privacy cost function through a contrastive self-supervised loss. Employing existing protocols of known-action and privacy attributes, our framework achieves a competitive action-privacy trade-off to the existing state-of-the-art supervised methods. In addition, we introduce a new protocol to evaluate the generalization of learned the anonymization function to novel-action and privacy at-tributes and show that our self-supervised framework out-performs existing supervised methods. Code available at: https://github.com/DAVEISHAN/SPAct

Federated learning (FL) is a promising strategy for performing privacy-preserving, distributed learning with a network of clients (i.e., edge devices). However, the data distribution among clients is often non-IID in nature, making efficient optimization difficult. To alleviate this issue, many FL algorithms focus on mitigating the effects of data heterogeneity across clients by introducing a variety of proximal terms, some incurring considerable compute and/or memory overheads, to restrain local updates with respect to the global model. Instead, we consider rethinking solutions to data heterogeneity in FL with a focus on local learning generality rather than proximal restriction. To this end, we first present a systematic study informed by second-order indicators to better understand algorithm effectiveness in FL. Interestingly, we find that standard regularization methods are surprisingly strong performers in mitigating data heterogeneity effects. Based on our findings, we further propose a simple and effective method, FedAlign, to overcome data heterogeneity and the pitfalls of previous methods. FedAlign achieves competitive accuracy with state-of-the-art FL methods across a variety of settings while minimizing computation and memory overhead. Code is available at https://github.com/mmendiet/FedAlign.

The dominant CNN-based methods for cross-view image geo-localization rely on polar transform and fail to model global correlation. We propose a pure transformer-based approach (TransGeo) to address these limitations from a different perspective. TransGeo takes full advantage of the strengths of transformer related to global information modeling and explicit position information encoding. We further leverage the flexibility of transformer input and propose an attention-guided non-uniform cropping method, so that uninformative image patches are removed with negligible drop on performance to reduce computation cost. The saved computation can be reallocated to increase resolution only for informative patches, resulting in performance improvement with no additional computation cost. This “attend and zoom-in” strategy is highly similar to human behavior when observing images. Remarkably, TransGeo achieves state-of-the-art results on both urban and rural datasets, with significantly less computation cost than CNN-based methods. It does not rely on polar transform and infers faster than CNN-based methods. Code is available at https://github.com/Jeff-Zilence/TransGeo2022.

Source code authorship attribution is an important problem often encountered in applications such as software forensics, bug fixing, and software quality analysis. Recent studies show that current source code authorship attribution methods can be compromised by attackers exploiting adversarial examples and coding style manipulation. This calls for robust solutions to the problem of code authorship attribution. In this paper, we initiate the study on making Deep Learning (DL)-based code authorship attribution robust. We propose an innovative framework called Robust coding style Patterns Generation (RoPGen), which essentially learns authors' unique coding style patterns that are hard for attackers to manipulate or imitate. The key idea is to combine data augmentation and gradient augmentation at the adversarial training phase. This effectively increases the diversity of training examples, generates meaningful perturbations to gradients of deep neural networks, and learns diversified representations of coding styles. We evaluate the effectiveness of RoPGen using four datasets of programs written in C, C++, and Java. Experimental results show that RoPGen can significantly improve the robustness of DL-based code authorship attribution, by respectively reducing 22.8% and 41.0% of the success rate of targeted and untargeted attacks on average.

Convolutional neural network (CNN) and Transformer have achieved great success in multimedia applications. However, little effort has been made to effectively and efficiently harmonize these two architectures to satisfy image deraining. This paper aims to unify these two architectures to take advantage of their learning merits for image deraining. In particular, the local connectivity and translation equivariance of CNN and the global aggregation ability of self-attention (SA) in Transformer are fully exploited for specific local context and global structure representations. Based on the observation that rain distribution reveals the degradation location and degree, we introduce degradation prior to help background recovery and accordingly present the association refinement deraining scheme. A novel multi-input attention module (MAM) is proposed to associate rain perturbation removal and background recovery. Moreover, we equip our model with effective depth-wise separable convolutions to learn the specific feature representations and trade off computational complexity. Extensive experiments show that our proposed method (dubbed as ELF) outperforms the state-of-the-art approach (MPRNet) by 0.25 dB on average, but only accounts for 11.7% and 42.1% of its computational cost and parameters.

Existing deep learning-based human mesh reconstruction approaches have a tendency to build larger networks to achieve higher accuracy. Computational complexity and model size are often neglected, despite being key characteristics for practical use of human mesh reconstruction models (e.g. virtual try-on systems). In this paper, we present GTRS, a lightweight pose-based method that can reconstruct human mesh from 2D human pose. We propose a pose analysis module that uses graph transformers to exploit structured and implicit joint correlations, and a mesh regression module that combines the extracted pose feature with the mesh template to reconstruct the final human mesh. We demonstrate the efficiency and generalization of GTRS by extensive evaluations on the Human3.6M and 3DPW datasets. In particular, GTRS achieves better accuracy than the SOTA pose-based method Pose2Mesh while only using 10.2% of the parameters (Params) and 2.5% of the FLOPs on the challenging in-the-wild 3DPW dataset. Code is available at https://github.com/zczcwh/GTRS

Fast and effective responses are required when a natural disaster (e.g., earthquake and hurricane) strikes. Building damage assessment from satellite imagery is critical before relief effort is deployed. With a pair of predisaster and postdisaster satellite images, building damage assessment aims at predicting the extent of damage to buildings. With the powerful ability of feature representation, deep neural networks have been successfully applied to building damage assessment. Most existing works simply concatenate predisaster and postdisaster images as input of a deep neural network without considering their correlations. In this article, we propose a novel two-stage convolutional neural network for building damage assessment, called BDANet. In the first stage, a U-Net is used to extract the locations of buildings. Then, the network weights from the first stage are shared in the second stage for building damage assessment. In the second stage, a two-branch multiscale U-Net is employed as the backbone, where predisaster and postdisaster images are fed into the network separately. A cross-directional attention module is proposed to explore the correlations between predisaster and postdisaster images. Moreover, CutMix data augmentation is exploited to tackle the challenge of difficult classes. The proposed method achieves state-of-the-art performance on a large-scale dataset—xBD. The code is available athttps://github.com/ShaneShen/BDANet-Building-Damage-Assessment.

AC series arc is dangerous and can cause serious electric fire hazards and property damage. This article proposed a convolutional neural network -based arc detection model named ArcNet. The database of this research is collected from eight different types of loads according to IEC62606 standard. The two most common types of arcs, including arcs from a loose connection of cables and those caused by the failure of the insulation, are generated in testing and included in the database. Using the database of raw current, experimental results indicate ArcNet can achieve a maximum of 99.47% arc detection accuracy at 10 kHz sampling rate. The model is also implemented in Raspberry Pi 3B for classification accuracy. A tradeoff study between the arc detection accuracy and model runtime has been conducted. The proposed ArcNet obtained an average runtime of 31 ms/sample of 1 cycle at 10 kHz sampling rate, which proves the feasibility of practical hardware deployment for real-time processing.

The encoder-decoder architecture is widely used as a lightweight semantic segmentation network. However, it struggles with a limited performance compared to a well-designed Dilated-FCN model for two major problems. First, commonly used upsampling methods in the decoder such as interpolation and deconvolution suffer from a local receptive field, unable to encode global contexts. Second, low-level features may bring noises to the network decoder through skip connections for the inadequacy of semantic concepts in early encoder layers. To tackle these challenges, a Global Enhancement Method is proposed to aggregate global information from high-level feature maps and adaptively distribute them to different decoder layers, alleviating the shortage of global contexts in the upsampling process. Besides, aLocal Refinement Module is developed by utilizing the decoder features as the semantic guidance to refine the noisy encoder features before the fusion of these two (the decoder features and the encoder features). Then, the two methods are integrated into a Context Fusion Block, and based on that, a novel Attention guided Global enhancement and Local refinement Network (AGLN) is elaborately designed. Extensive experiments on PASCAL Context, ADE20K, and PASCAL VOC 2012 datasets have demonstrated the effectiveness of the proposed approach. In particular, with a vanilla ResNet-101 backbone, AGLN achieves the state-of-the-art result (56.23% mean IOU) on the PASCAL Context dataset. The code is available at https://github.com/zhasen1996/AGLN.

The papers in this special section focus on emerging Internet of Medical Things. Recent advances in advances in healthcare can be experienced with the development of smart sensorial things, Artificial Intelligence (AI), Machine Learning (ML), Deep Learning (DL), edge computing, Edge AI, 6G, cloud computing, and connected healthcare have attracted a great deal of attention and a wide range of views. However, the need to deliver real-time and accurate healthcare services to patients, while reducing costs is a challenging issue [1]. Especially, COVID-19 has recently demonstrated the importance of fast, comprehensive, and accurate intelligent healthcare involving different types of medical, physiological, and epidemiological investigation data to diagnose the virus. Smart health is a real-time, intelligent, ubiquitous healthcare service based on Internet of bioMedical Things (IoMT). With the rapid development of related technologies such as deep learning, edge computing and IoT, smart health is playing vital role in healthcare industry to increase the accuracy, reliability, and productivity of mobile sensory devices.

In the United States, highway-railroad grade crossings are easily congested, which not only causes significant traffic delays to travelers but also brings potential threats to the first responders for emergencies. Unfortunately, very limited research efforts have been dedicated to developing practical systems that can assess traffic conditions at overcrowded grade crossings. The main challenge in evaluating the congestion conditions at the crossings is the different instance classes (i.e., vehicle, train, and pedestrian) that need to be accurately detected, especially when densely packaged. In this study, a novel convolutional neural network (CNN) named dense traffic detection net (DTDNet) is developed. DTDNet proposes to integrate the Transformer Attention (TA) module for better modeling of global context information and the learning-to-match detection head for optimizing object detection and localization using a likelihood probability fashion. To train and test DTDNet, a unique grade crossing traffic image dataset including congested and normal traffic during both daytime and nighttime is established. Experimental results on the dataset show that the proposed DTDNet achieves the maximum mean average precision (mAP) value, 0.832, outperforming the other state-of-the-art (SOTA) models. Field test results with low mean average error (MAE), mean relative error (MRE), and root mean squared error (RMSE) which are 2.200, 1.890, and 0.280, respectively suggest the proposed model has a satisfying and robust performance in the field application under different environments.”.

Frame reconstruction (current or future frame) based on Auto-Encoder (AE) is a popular method for video anomaly detection. With models trained on the normal data, the reconstruction errors of anomalous scenes are usually much larger than those of normal ones. Previous methods introduced the memory bank into AE, for encoding diverse normal patterns across the training videos. However, they are memory-consuming and cannot cope with unseen new scenarios in the testing data. In this work, we propose a dynamic prototype unit (DPU) to encode the normal dynamics as prototypes in real time, free from extra memory cost. In addition, we introduce meta-learning to our DPU to form a novel few-shot normalcy learner, namely Meta-Prototype Unit (MPU). It enables the fast adaption capability on new scenes by only consuming a few iterations of update. Extensive experiments are conducted on various benchmarks. The superior performance over the state-of-the-art demonstrates the effectiveness of our method. Our code is available at https://github.com/ktr-hubrt/MPN/.

Cross-view image geo-localization aims to determine the locations of street-view query images by matching with GPS-tagged reference images from aerial view. Recent works have achieved surprisingly high retrieval accuracy on city-scale datasets. However, these results rely on the assumption that there exists a reference image exactly centered at the location of any query image, which is not applicable for practical scenarios. In this paper, we redefine this problem with a more realistic assumption that the query image can be arbitrary in the area of interest and the reference images are captured before the queries emerge. This assumption breaks the one-to-one retrieval setting of existing datasets as the queries and reference images are not perfectly aligned pairs, and there may be multiple reference images covering one query location. To bridge the gap between this realistic setting and existing datasets, we propose a new large-scale benchmark –VIGOR– for cross-View Image Geo-localization beyond One-to-one Retrieval. We benchmark existing state-of-the-art methods and propose a novel end-to-end framework to localize the query in a coarse-to-fine manner. Apart from the image-level retrieval accuracy, we also evaluate the localization accuracy in terms of the actual distance (meters) using the raw GPS data. Extensive experiments are conducted under different application scenarios to validate the effectiveness of the proposed method. The results indicate that cross-view geo-localization in this realistic setting is still challenging, fostering new research in this direction. Our dataset and code will be released at https://github.com/JeffZilence/VIGOR.

Transformer architectures have become the model of choice in natural language processing and are now being introduced into computer vision tasks such as image classification, object detection, and semantic segmentation. However, in the field of human pose estimation, convolutional architectures still remain dominant. In this work, we present PoseFormer, a purely transformer-based approach for 3D human pose estimation in videos without convolutional architectures involved. Inspired by recent developments in vision transformers, we design a spatial-temporal transformer structure to comprehensively model the human joint relations within each frame as well as the temporal correlations across frames, then output an accurate 3D human pose of the center frame. We quantitatively and qualitatively evaluate our method on two popular and standard benchmark datasets: Human3.6M and MPI-INF-3DHP. Extensive experiments show that PoseFormer achieves state-of-the-art performance on both datasets. Code is available at https://github.com/zczcwh/PoseFormer

Image deraining is an effective solution to avoid performance drop of vision-oriented tasks in rainy weather. Most existing image deraining approaches either fail to produce satisfactory restoration results or cost too much computation. In this paper, we propose a low-complexity and high-performance coupled representation module (CRM), designed to learn the joint features of rain-free contents and rain information as well as their blending correlations. To promote the computation efficiency, we employ depth-wise separable convolutions, and construct CRM in an asymmetric U-shaped architecture to reduce model parameters and memory footprint. Our final model–PCNet achieves the progressive separation of rain-free contents and rain streaks using cascaded residual learning. Extensive experiments are conducted to evaluate the efficacy of the proposed PCNet on several synthetic and real-world rain datasets.

Domain adaptive image retrieval (DAR) aims to train the model with well-labeled source domain and target images in order to retrieve source instances given query target samples from the identical category space. However, the practical scenario hinders to manually annotate all retrieved images due to huge labeling cost. Motivated by the realistic demand, we firstly define the semi-supervised domain adaptive retrieval (SDAR) problem, assuming the database includes a small proportion annotated source images and abundant unlabeled ones. To overcome the challenging SDAR, this paper propose a novel method named Discriminative Hashing learning (DHLing) which mainly includes two modules, i.e., domain-specific optimization and domain-invariant memory bank. Specifically, the first component explores the structural knowledge of samples to predict the unlabeled images with pseudo labels to achieve hash coding consistency. While, the second one attempts to construct the domain-invariant memory bank to guide the feature generation and achieve cross-domain alignment. Experimental results on several popular cross-domain retrieval benchmarks illustrate the effectiveness of our proposed DHLing on both conventional DAR and new SDAR scenarios by comparing with the state-of-the-art retrieval methods.

Existing methods for object detection in UAV images ignored an important challenge - imbalanced class distribution in UAV images - which leads to poor performance on tail classes. We systematically investigate existing solutions to long-tail problems and unveil that re-balancing methods that are effective on natural image datasets cannot be trivially applied to UAV datasets. To this end, we rethink longtailed object detection in UAV images and propose the Dual Sampler and Head detection Network (DSHNet), which is the first work that aims to resolve long-tail distribution in UAV images. The key components in DSHNet include Class-Biased Samplers (CBS) and Bilateral Box Heads (BBH), which are developed to cope with tail classes and head classes in a dual-path manner. Without bells and whistles, DSHNet significantly boosts the performance of tail classes on different detection frameworks. Moreover, DSHNet significantly outperforms base detectors and generic approaches for long-tail problems on VisDrone and UAVDT datasets. It achieves new state-of-the-art performance when combining with image cropping methods. Code is available at https://github.com/we1pingyu/DSHNet.

Street-to-aerial image geo-localization, which matches a query street-view image to the GPS-tagged aerial images in a reference set, has attracted increasing attention recently. In this paper, we revisit this problem and point out the ignored issue about image alignment information. We show that the performance of a simple Siamese network is highly dependent on the alignment setting and the comparison of previous works can be unfair if they have different assumptions. Instead of focusing on the feature extraction under the alignment assumption, we show that improvements in metric learning techniques significantly boost the performance regardless of the alignment. Without leveraging the alignment information, our pipeline outperforms previous works on both panorama and cropped datasets. Furthermore, we conduct visualization to help understand the learned model and the effect of alignment information using Grad-CAM. With our discovery on the approximate rotation-invariant activation maps, we propose a novel method to estimate the orientation/alignment between a pair of cross-view images with unknown alignment information. It achieves state-of-the-art results on the CVUSA dataset.

Rain streaks in the air show diverse characteristics with different shapes, directions, densities, even the complex overlapped phenomenon, causing great challenges for the deraining task. Recently, deep learning based image deraining methods have been extensively investigated due to their excellent performance. However, most of the existing algorithms still have limitations in removing rain streaks while preserving rich textural details under complicated rain conditions. To this end, we propose to decompose rain streaks into multiple rain layers and individually estimate each of them along the network stages to cope with the increasing abstracts. To better characterize rain layers, an improved non-local block is designed to exploit the self-similarity of rain information by learning the holistic spatial feature correlations while reducing the calculation complexity. Moreover, a mixed attention mechanism is applied to guide the fusion of rain layers by focusing on the local and global overlaps among these rain layers. Extensive experiments on both synthetic rainy/rain-haze/raindrop datasets, real-world samples, the haze, and low-light scenarios show substantial improvements both on quantitative indicators and visual effects over the current state-of-the-art technologies. The source code is available athttps://github.com/kuihua/IADN.

Deep-learning-based methods, such as convolution neural network (CNN), have demonstrated their efficiency in hyperspectral image (HSI) classification. These methods can automatically learn spectral-spatial discriminative features within local patches. However, for each pixel in an HSI, it is not only related to its nearby pixels but also has connections to pixels far away from itself. Therefore, to incorporate the long-range contextual information, a deep fully convolutional network (FCN) with an efficient nonlocal module, named ENL-FCN, is proposed for HSI classification. In the proposed framework, a deep FCN considers an entire HSI as input and extracts spectral-spatial information in a local receptive field. The efficient nonlocal module is embedded in the network as a learning unit to capture the long-range contextual information. Different from the traditional nonlocal neural networks, the long-range contextual information is extracted in a specially designed criss-cross path for computation efficiency. Furthermore, using a recurrent operation, each pixel's response is aggregated from all pixels of HSI. The benefits of our proposed ENL-FCN are threefold: 1) the long-range contextual information is incorporated effectively; 2) the efficient module can be freely embedded in a deep neural network in a plug-and-play fashion; and 3) it has much fewer learning parameters and requires less computational resources. The experiments conducted on three popular HSI data sets demonstrate that the proposed method achieves state-of-the-art classification performance with lower computational cost in comparison with several leading deep neural networks for HSI.

Rainy weather is a challenge for many vision-oriented tasks (e.g., object detection and segmentation), which causes performance degradation. Image deraining is an effective solution to avoid performance drop of downstream vision tasks. However, most existing deraining methods either fail to produce satisfactory restoration results or cost too much computation. In this work, considering both effectiveness and efficiency of image deraining, we propose a progressive coupled network (PCNet) to well separate rain streaks while preserving rain-free details. To this end, we investigate the blending correlations between them and particularly devise a novel coupled representation module (CRM) to learn the joint features and the blending correlations. By cascading multiple CRMs, PCNet extracts the hierarchical features of multi-scale rain streaks, and separates the rain-free content and rain streaks progressively. To promote computation efficiency, we employ depth-wise separable convolutions and a U-shaped structure, and construct CRM in an asymmetric architecture to reduce model parameters and memory footprint. Extensive experiments are conducted to evaluate the efficacy of the proposed PCNet in two aspects: (1) image deraining on several synthetic and real-world rain datasets and (2) joint image deraining and downstream vision tasks (e.g., object detection and segmentation). Furthermore, we show that the proposed CRM can be easily adopted to similar image restoration tasks including image dehazing and low-light enhancement with competitive performance. The source code is available at https://github.com/kuijiang0802/PCNet.

This work explores the visual explanation for deep metric learning and its applications. As an important problem for learning representation, metric learning has attracted much attention recently, while the interpretation of the metric learning model is not as well-studied as classification. To this end, we propose an intuitive idea to show where contributes the most to the overall similarity of two input images by decomposing the final activation. Instead of only providing the overall activation map of each image, we propose to generate point-to-point activation intensity between two images so that the relationship between different regions is uncovered. We show that the proposed framework can be directly applied to a wide range of metric learning applications and provides valuable information for model understanding. Both theoretical and empirical analyses are provided to demonstrate the superiority of the proposed overall activation map over existing methods. Furthermore, our experiments validate the effectiveness of the proposed point-specific activation map on two applications, i.e. cross-view pattern discovery and interactive retrieval. Code is available at https://github.com/Jeff-Zilence/Explain_Metric_Learning.

Rain streaks in the air appear in various blurring degrees and resolutions due to different distances from their positions to the camera. Similar rain patterns are visible in a rain image as well as its multi-scale (or multi-resolution) versions, which makes it possible to exploit such complementary information for rain streak representation. In this work, we explore the multi-scale collaborative representation for rain streaks from the perspective of input image scales and hierarchical deep features in a unified framework, termed multi-scale progressive fusion network (MSPFN) for single image rain streak removal. For the similar rain streaks at different positions, we employ recurrent calculation to capture the global texture, thus allowing to explore the complementary and redundant information at the spatial dimension to characterize target rain streaks. Besides, we construct multi-scale pyramid structure, and further introduce the attention mechanism to guide the fine fusion of these correlated information from different scales. This multi-scale progressive fusion strategy not only promotes the cooperative representation, but also boosts the end-to-end training. Our proposed method is extensively evaluated on several benchmark datasets and achieves the state-of-the-art results. Moreover, we conduct experiments on joint deraining, detection, and segmentation tasks, and inspire a new research direction of vision task driven image deraining. The source code is available at https://github.com/kuihua/MSPFN.

We propose a novel attention-based framework for 3D human pose estimation from a monocular video. Despite the general success of end-to-end deep learning paradigms, our approach is based on two key observations: (1) temporal incoherence and jitter are often yielded from a single frame prediction; (2) error rate can be remarkably reduced by increasing the receptive field in a video. Therefore, we design an attentional mechanism to adaptively identify significant frames and tensor outputs from each deep neural net layer, leading to a more optimal estimation. To achieve large temporal receptive fields, multi-scale dilated convolutions are employed to model long-range dependencies among frames. The architecture is straightforward to implement and can be flexibly adopted for real-time applications. Any off-the-shelf 2D pose estimation system, e.g. Mocap libraries, can be easily integrated in an ad-hoc fashion. We both quantitatively and qualitatively evaluate our method on various standard benchmark datasets (e.g. Human3.6M, HumanEva). Our method considerably outperforms all the state-of-the-art algorithms up to 8% error reduction (average mean per joint position error: 34.7) as compared to the best-reported results. Code is available at: (https://github.com/lrxjason/Attention3DHumanPose)

Due to diverse rain shapes, directions, densities as well as different distances to cameras, rain streaks in the air are interweaved and overlapped. However, most existing deraining methods are inherently oblivious this phenomenon and tend to learn a single rain streak layer to simulate this complex distribution, consequently failing to restore high-quality rain-free images. To solve this problem, along with the stage-wise learning, we propose a novel attention-guided deraining network (ADN) for rain streak removal. Specially, we decompose the rain streaks into multiple rain streak layers, and individually model them along the stages of the network to match the increasing abstracts. Moreover, the attention mechanism is utilized to guide the fusion of these rain streak layers by handling the overlaps between them. Extensive experiments on several benchmark datasets and real-world scenarios show substantial improvements both on quantitative indicators and visual effects over the current top-performing methods.

Neural architecture search (NAS) proves to be among the best approaches for many tasks by generating an application-adaptive neural architectures, which are still challenged by high computational cost and memory consumption. At the same time, 1-bit convolutional neural networks (CNNs) with binarized weights and activations show their potential for resource-limited embedded devices. One natural approach is to use 1-bit CNNs to reduce the computation and memory cost of NAS by taking advantage of the strengths of each in a unified framework. To this end, a Child-Parent model is introduced to a differentiable NAS to search the binarized architecture(Child) under the supervision of a full-precision model (Parent). In the search stage, the Child-Parent model uses an indicator generated by the parent and child model accuracy to evaluate the performance and abandon operations with less potential. In the training stage, a kernel level CP loss is introduced to optimize the binarized network. Extensive experiments demonstrate that the proposed CP-NAS achieves a comparable accuracy with traditional NAS on both the CIFAR and ImageNet databases. It achieves an accuracy of 95.27% on CIFAR-10, 64.3% on ImageNet with binarized weights and activations, and a 30% faster search than prior arts.

We propose a new regularization method to alleviate over-fitting in deep neural networks. The key idea is utilizing randomly transformed training samples to regularize a set of sub-networks, which are originated by sampling the width of the original network, in the training process. As such, the proposed method introduces self-guided disturbances to the raw gradients of the network and therefore is termed as Gradient Augmentation (GradAug). We demonstrate that GradAug can help the network learn well-generalized and more diverse representations. Moreover, it is easy to implement and can be applied to various structures and applications. GradAug improves ResNet-50 to 78.79% on ImageNet classification, which is a new state-of-the-art accuracy. By combining with CutMix, it further boosts the performance to 79.67%, which outperforms an ensemble of advanced training tricks. The generalization ability is evaluated on COCO object detection and instance segmentation where GradAug significantly surpasses other state-of-the-art methods. GradAug is also robust to image distortions and FGSM adversarial attacks and is highly effective in low data regimes. Code is available at \url{https://github.com/taoyang1122/GradAug}

Pedestrian detection has remained an important research topic in both the computer vision and multimedia communities because of its importance in practical applications, such as driving assistance and video surveillance. Existing methods compare the response score with a fixed threshold to determine whether a candidate region contains pedestrians and produce dissatisfactory results that contain either missed detections or false detections, which are difficult to balance. This situation has a serious impact under the condition of variable scale. This paper investigates the functional relationship between the scores and scales of pedestrians. By designing experiments with multiple scales, we have found a discriminant surface in the score scale space. Pedestrians can be distinguished at various scale levels according to their locations on the discriminant surface. The proposed approach is evaluated using four challenging pedestrian detection datasets, including Caltech, INRIA, ETH, and KITTI, and the superior experimental results are achieved when compared with baseline methods.

Human action recognition aims to classify a given video according to which type of action it contains. Disturbance brought by clutter background and unrelated motions makes the task challenging for video frame-based methods. To solve this problem, this paper takes advantage of pose estimation to enhance the performances of video frame features. First, we present a pose feature called dynamic pose image (DPI), which describes human action as the aggregation of a sequence of joint estimation maps. Different from traditional pose features using sole joints, DPI suffers less from disturbance and provides richer information about human body shape and movements. Second, we present attention-based dynamic texture images (att-DTIs) as pose-guided video frame feature. Specifically, a video is treated as a space-time volume, and DTIs are obtained by observing the volume from different views. To alleviate the effect of disturbance on DTIs, we accumulate joint estimation maps as attention map, and extend DTIs to attention-based DTIs (att-DTIs). Finally, we fuse DPI and att-DTIs with multi-stream deep neural networks and late fusion scheme for action recognition. Experiments on NTU RGB+D, UTD-MHAD, and Penn-Action datasets show the effectiveness of DPI and att-DTIs, as well as the complementary property between them.

In stochastic gradient descent (SGD) and its variants, the optimized gradient estimators may be as expensive to compute as the true gradient in many scenarios. This paper introduces a calibrated stochastic gradient descent (CSGD) algorithm for deep neural network optimization. A theorem is developed to prove that an unbiased estimator for the network variables can be obtained in a probabilistic way based on the Lipschitz hypothesis. Our work is significantly distinct from existing gradient optimization methods, by providing a theoretical framework for unbiased variable estimation in the deep learning paradigm to optimize the model parameter calculation. In particular, we develop a generic gradient calibration layer which can be easily used to build convolutional neural networks (CNNs). Experimental results demonstrate that CNNs with our CSGD optimization scheme can improve the stateof-the-art performance for natural image classification, digit recognition, ImageNet object classification, and object detection tasks. This work opens new research directions for developing more efficient SGD updates and analyzing the backpropagation algorithm.

The task of cross-view image geo-localization aims to determine the geo-location (GPS coordinates) of a query ground-view image by matching it with the GPS-tagged aerial (satellite) images in a reference dataset. Due to the dramatic changes of viewpoint, matching the cross-view images is challenging. In this paper, we propose the GeoCapsNet based on the capsule network for ground-to-aerial image geo-localization. The network first extracts features from both ground and aerial images via standard convolution layers and the capsule layers further encode the features to model the spatial feature hierarchies and enhance the representation power. Moreover, we introduce a simple and effective weighted soft-margin triplet loss with online batch hard sample mining, which can greatly improve the image retrieval accuracy. Experimental results show that our GeoCapsNet significantly outperforms the state-of-the-art approaches on two benchmark datasets.

Convolutional neural network (CNN) receives extensive attention in hyperspectral image classification. While hyper-spectral images contain abundant spectral information but lack spatial information, which usually contributes to poor classification results. In this paper, a novel classification framework called information fusion based CNN (IF-CNN) is proposed to compensate for the shortcomings of hyper-spectral images. The proposed method merges hyperspectral images with abundant spectral information and LiDAR images with rich spatial information as the input of classification framework. Furthermore, the framework consists of two convolutional neural networks: one-dimensional CNN for extracting spectral features, and two-dimensional CNN for extracting spatial correlation features. Experimental results demonstrate that the proposed method achieves excellent performance compared with some existing methods.

Discriminative local features obtained from activations of convolutional neural networks have proven to be essential for image retrieval. To improve retrieval performance, many recent works aim to obtain more powerful and discriminative features. In this work, we propose a new attention layer to assess the importance of local features and assign higher weights to those more discriminative. Furthermore, we present a scale and mask module to filter out the meaningless local features and scale the major components. This module not only reduces the impact of the various scales of the major components in images by scaling them on the feature maps, but also filters out the redundant and confusing features with the MAX-Mask. Finally, the features are aggregated into the image representation. Experimental evaluations demonstrate that the proposed method outperforms the state-of-the-art methods on standard image retrieval datasets.

In this study, the authors focus on the challenging problem of verifying faces captured under unconstrained conditions. Unconstrained face images often vary largely in poses, illuminations, expressions, occlusions, and ages. To address these challenges, they combine face frontalisation method with metric learning. To deal with the variations of poses, they apply an improved 3D face frontalisation method to generate the frontal view of the face images. Recent studies observed that bilinear similarity and Mahalanobis distance have a promising performance on measuring the similarity of two images. Based on these studies, they propose a weighted similarity and distance metric learning method which balances the role of bilinear similarity and Mahalanobis distance to better measure the similarity of an image pair. All the experiments are conducted based on the labelled faces in the wild database, and the experimental results show the effectiveness of their method.

Convolutional neural network (CNN) has been widely used in hyperspectral imagery (HSI) classification. Data augmentation is proven to be quite effective when training data size is relatively small. In this letter, extensive comparison experiments are conducted with common data augmentation methods, which draw an observation that common methods can produce a limited and up-bounded performance. To address this problem, a new data augmentation method, named as pixel-block pair (PBP), is proposed to greatly increase the number of training samples. The proposed method takes advantage of deep CNN to extract PBP features, and decision fusion is utilized for final label assignment. Experimental results demonstrate that the proposed method can outperform the existing ones.

Label information plays an important role in a supervised hyperspectral image classification problem. However, current classification methods all ignore an important and inevitable problem-labels may be corrupted and collecting clean labels for training samples is difficult and often impractical. Therefore, how to learn from the database with noisy labels is a problem of great practical importance. In this paper, we study the influence of label noise on hyperspectral image classification and develop a random label propagation algorithm (RLPA) to cleanse the label noise. The key idea of RLPA is to exploit knowledge (e.g., the superpixel-based spectral-spatial constraints) from the observed hyperspectral images and apply it to the process of label propagation. Specifically, the RLPA first constructs a spectral-spatial probability transform matrix (SSPTM) that simultaneously considers the spectral similarity and superpixel-based spatial information. It then randomly chooses some training samples as “clean” samples and sets the rest as unlabeled samples, and propagates the label information from the “clean” samples to the rest unlabeled samples with the SSPTM. By repeating the random assignment (of “clean” labeled samples and unlabeled samples) and propagation, we can obtain multiple labels for each training sample. Therefore, the final propagated label can be calculated by a majority vote algorithm. Experimental studies show that the RLPA can reduce the level of noisy label and demonstrates the advantages of our proposed method over four major classifiers with a significant margin-the gains in terms of the average overall accuracy, average accuracy, and kappa are impressive, e.g., 9.18%, 9.58%, and 0.1043. The MATLAB source code is available at https://github.com/junjun-jiang/RLPA.

While intrinsic data structure in subspace provides useful information for visual recognition, it has not yet been well studied in deep feature learning for action recognition. In this paper, we introduce a new spatio-temporal manifold network (STMN) that leverages data manifold structures to regularize deep action feature learning, aiming at simultaneously minimizing the intra-class variations of learned deep features and alleviating the over-fitting problem. To this end, the manifold prior is imposed from the top layer of a convolutional neural network (CNN), and is propagated across convolutional layers during forward-backward propagation. The observed correspondence of manifold structures in the data space and feature space validates that the manifold priori can be transferred across CNN layers. STMN theoretically recasts the problem of transferring the data structure prior into the deep learning architectures as a projection over the manifold via an embedding method, which can be easily solved by an Alternating Direction Method of Multipliers and Backward Propagation (ADMM-BP) algorithm. STMN is generic in the sense that it can be plugged into various backbone architectures to learn more discriminative representation for action recognition. Extensive experimental results show that our method achieves comparable or even better performance as compared with the state-of-the-art approaches on four benchmark datasets.

Although recent semisupervised approaches have proven their effectiveness when there are limited training data, they assume that the samples from different actions lie on a single data manifold in the feature space and try to uncover a common subspace for all samples. However, this assumption ignores the intraclass compactness and the interclass separability simultaneously. We believe that human actions should occupy multimanifold subspace and, therefore, model the samples of the same action as the same manifold and those of different actions as different manifolds. In order to obtain the optimum subspace projection matrix, the current approaches may be mathematically imprecise owe to the badly scaled matrix and improper convergence. To address these issues in unconstrained convex optimization, we introduce a nontrivial spectral projected gradient method and Karush-Kuhn-Tucker conditions without matrix inversion. Through maximizing the separability between different classes by using labeled data points and estimating the intrinsic geometric structure of the data distributions by exploring unlabeled data points, the proposed algorithm can learn global and local consistency and boost the recognition performance. Extensive experiments conducted on the realistic video data sets, including JHMDB, HMDB51, UCF50, and UCF101, have demonstrated that our algorithm outperforms the compared algorithms, including deep learning approach when there are only a few labeled samples.

Surveillance videos are able to capture a variety of realistic anomalies. In this paper, we propose to learn anomalies by exploiting both normal and anomalous videos. To avoid annotating the anomalous segments or clips in training videos, which is very time consuming, we propose to learn anomaly through the deep multiple instance ranking framework by leveraging weakly labeled training videos, i.e. the training labels (anomalous or normal) are at video-level instead of clip-level. In our approach, we consider normal and anomalous videos as bags and video segments as instances in multiple instance learning (MIL), and automatically learn a deep anomaly ranking model that predicts high anomaly scores for anomalous video segments. Furthermore, we introduce sparsity and temporal smoothness constraints in the ranking loss function to better localize anomaly during training. We also introduce a new large-scale first of its kind dataset of 128 hours of videos. It consists of 1900 long and untrimmed real-world surveillance videos, with 13 realistic anomalies such as fighting, road accident, burglary, robbery, etc. as well as normal activities. This dataset can be used for two tasks. First, general anomaly detection considering all anomalies in one group and all normal activities in another group. Second, for recognizing each of 13 anomalous activities. Our experimental results show that our MIL method for anomaly detection achieves significant improvement on anomaly detection performance as compared to the state-of-the-art approaches. We provide the results of several recent deep learning baselines on anomalous activity recognition. The low recognition performance of these baselines reveals that our dataset is very challenging and opens more opportunities for future work. The dataset is available at: http://crcv.ucf.edu/projects/real-world.

Skeleton-based action recognition task is entangled with complex spatio-temporal variations of skeleton joints, and remains challenging for Recurrent Neural Networks (RNNs). In this work, we propose a temporal-then-spatial recalibration scheme to alleviate such complex variations, resulting in an end-to-end Memory Attention Networks (MANs) which consist of a Temporal Attention Recalibration Module (TARM) and a Spatio-Temporal Convolution Module (STCM). Specifically, the TARM is deployed in a residual learning module that employs a novel attention learning network to recalibrate the temporal attention of frames in a skeleton sequence. The STCM treats the attention calibrated skeleton joint sequences as images and leverages the Convolution Neural Networks (CNNs) to further model the spatial and temporal information of skeleton data. These two modules (TARM and STCM) seamlessly form a single network architecture that can be trained in an end-to-end fashion. MANs significantly boost the performance of skeleton-based action recognition and achieve the best results on four challenging benchmark datasets: NTU RGB+D, HDM05, SYSU-3D and UT-Kinect.

Steerable properties dominate the design of traditional filters, e.g., Gabor filters, and endow features the capability of dealing with spatial transformations. However, such excellent properties have not been well explored in the popular deep convolutional neural networks (DCNNs). In this paper, we propose a new deep model, termed Gabor Convolutional Networks (GCNs or Gabor CNNs), which incorporates Gabor filters into DCNNs to enhance the resistance of deep learned features to the orientation and scale changes. By only manipulating the basic element of DCNNs based on Gabor filters, i.e., the convolution operator, GCNs can be easily implemented and are compatible with any popular deep learning architecture. Experimental results demonstrate the super capability of our algorithm in recognizing objects, where the scale and rotation changes occur frequently. The proposed GCNs have much fewer learnable network parameters, and thus is easier to train with an endtoend pipeline. The source code will be here1.

This paper presents an effective multiscale energy-based global ternary image (E-GTI) representation for action recognition from depth sequences. The unique property of our representation is that it takes the spatiotemporal discrimination and action speed variations into account, intending to solve the problems of distinguishing similar actions and identifying the actions with different speeds in one goal. The entire method is carried out in two stages. In the first stage, consecutive depth frames are used to generate global ternary image (GTI) features, which implicitly capture both inter-frame motion regions and motion directions. Specifically, each pixel in the GTI represents one of three possible states, namely, positive, negative, and neutral, which indicate the increased, decreased, and same depth values, respectively. To cope with speed variations in actions, energy-based sampling method is utilized, leading to multiscale E-GTI features, where the multiscale scheme can efficiently capture the temporal relationships among frames. In the second stage, all the E-GTI features are transformed by Radon transform (RT) as robust descriptors, which are aggregated by the bag-of-visual-words model as a compact representation. Extensive experiments on benchmark data sets show that our representation outperforms state-of-the-art approaches, since it captures discriminating spatiotemporal information of actions. Due to the merits of energy-based sampling and RT methods, our representation shows robustness to speed variations, depth noise, and partial occlusions.

Person reidentification (re-id), as an important task in video surveillance and forensics applications, has been widely studied. Previous research efforts toward solving the person re-id problem have primarily focused on constructing robust vector description by exploiting appearance's characteristic, or learning discriminative distance metric by labeled vectors. Based on the cognition and identification process of human, we propose a new pattern, which transforms the feature description from characteristic vector to discrepancy matrix. In particular, in order to well identify a person, it converts the distance metric from vector metric to matrix metric, which consists of the intradiscrepancy projection and interdiscrepancy projection parts. We introduce a consistent term and a discriminative term to form the objective function. To solve it efficiently, we utilize a simple gradient-descent method under the alternating optimization process with respect to the two projections. Experimental results on public datasets demonstrate the effectiveness of the proposed pattern as compared with the state-of-the-art approaches.

As an unsupervised dimensionality reduction method, the principal component analysis (PCA) has been widely considered as an efficient and effective preprocessing step for hyperspectral image (HSI) processing and analysis tasks. It takes each band as a whole and globally extracts the most representative bands. However, different homogeneous regions correspond to different objects, whose spectral features are diverse. Therefore, it is inappropriate to carry out dimensionality reduction through a unified projection for an entire HSI. In this paper, a simple but very effective superpixelwise PCA (SuperPCA) approach is proposed to learn the intrinsic low-dimensional features of HSIs. In contrast to classical PCA models, the SuperPCA has four main properties: 1) unlike the traditional PCA method based on a whole image, the SuperPCA takes into account the diversity in different homogeneous regions, that is, different regions should have different projections; 2) most of the conventional feature extraction models cannot directly use the spatial information of HSIs, while the SuperPCA is able to incorporate the spatial context information into the unsupervised dimensionality reduction by superpixel segmentation; 3) since the regions obtained by superpixel segmentation have homogeneity, the SuperPCA can extract potential low-dimensional features even under noise; and 4) although the SuperPCA is an unsupervised method, it can achieve a competitive performance when compared with supervised approaches. The resulting features are discriminative, compact, and noise-resistant, leading to an improved HSI classification performance. Experiments on three public data sets demonstrate that the SuperPCA model significantly outperforms the conventional PCA-based dimensionality reduction baselines for HSI classification, and some state-of-the-art feature extraction approaches. The MATLAB source code is available at https://github.com/junjun-jiang/SuperPCA.

Satellite image denoising is essential for enhancing the visual quality of images and for facilitating further image processing and analysis tasks. Designing of self-tunable 2-D finite-impulse response (FIR) filters attracted researchers to explore its usefulness in various domains. Furthermore, 2-D FIR Wiener filters which estimate the desired signal using its statistical parameters became a standard method employed for signal restoration applications. In this paper, we propose a 2-D FIR Wiener filter driven by the adaptive cuckoo search (ACS) algorithm for denoising multispectral satellite images contaminated with the Gaussian noise of different variance levels. The ACS algorithm is proposed to optimize the Wiener weights for obtaining the best possible estimate of the desired uncorrupted image. Quantitative and qualitative comparisons are conducted with 10 recent denoising algorithms prominently used in the remote-sensing domain to substantiate the performance and computational capability of the proposed ACSWF. The tested data set included satellite images procured from various sources, such as Satpalda Geospatial Services, Satellite Imaging Corporation, and National Aeronautics and Space Administration. The stability analysis and study of convergence characteristics are also performed, which revealed the possibility of extending the ACSWF for real-time applications as well.

In steerable filters, a filter of arbitrary orientation can be generated by a linear combination of a set of "basis filters." Steerable properties dominate the design of the traditional filters, e.g., Gabor filters and endow features the capability of handling spatial transformations. However, such properties have not yet been well explored in the deep convolutional neural networks (DCNNs). In this paper, we develop a new deep model, namely, Gabor convolutional networks (GCNs or Gabor CNNs), with Gabor filters incorporated into DCNNs such that the robustness of learned features against the orientation and scale changes can be reinforced. By manipulating the basic element of DCNNs, i.e., the convolution operator, based on Gabor filters, GCNs can be easily implemented and are readily compatible with any popular deep learning architecture. We carry out extensive experiments to demonstrate the promising performance of our GCNs framework, and the results show its superiority in recognizing objects, especially when the scale and rotation changes take place frequently. Moreover, the proposed GCNs have much fewer network parameters to be learned and can effectively reduce the training complexity of the network, leading to a more compact deep learning model while still maintaining a high feature representation capacity. The source code can be found at https://github.com/bczhangbczhang.

Correlation filters are special classifiers designed for shift-invariant object recognition, which are robust to pattern distortions. The recent literature shows that combining a set of sub-filters trained based on a single or a small group of images obtains the best performance. The idea is equivalent to estimating variable distribution based on the data sampling (bagging), which can be interpreted as finding solutions (variable distribution approximation) directly from sampled data space. However, this methodology fails to account for the variations existed in the data. In this paper, we introduce an intermediate step-solution sampling-after the data sampling step to form a subspace, in which an optimal solution can be estimated. More specifically, we propose a new method, named latent constrained correlation filters (LCCF), by mapping the correlation filters to a given latent subspace, and develop a new learning framework in the latent subspace that embeds distribution-related constraints into the original problem. To solve the optimization problem, we introduce a subspace-based alternating direction method of multipliers, which is proven to converge at the saddle point. Our approach is successfully applied to three different tasks, including eye localization, car detection, and object tracking. Extensive experiments demonstrate that LCCF outperforms the state-of-the-art methods.11.

Three-dimensional (3-D) action recognition has broad applications in human-computer interaction and intelligent surveillance. However, recognizing similar actions remains challenging since previous literature fails to capture motion and shape cues effectively from noisy depth data. In this paper, we propose a novel two-layer Bag-of-Visual-Words (BoVW) model, which suppresses the noise disturbances and jointly encodes both motion and shape cues. First, background clutter is removed by a background modeling method that is designed for depth data. Then, motion and shape cues are jointly used to generate robust and distinctive spatial-temporal interest points (STIPs): motion-based STIPs and shape-based STIPs. In the first layer of our model, a multiscale 3-D local steering kernel descriptor is proposed to describe local appearances of cuboids around motion-based STIPs. In the second layer, a spatial-temporal vector descriptor is proposed to describe the spatial-temporal distributions of shape-based STIPs. Using the BoVW model, motion and shape cues are combined to form a fused action representation. Our model performs favorably compared with common STIP detection and description methods. Thorough experiments verify that our model is effective in distinguishing similar actions and robust to background clutter, partial occlusions and pepper noise.

Traditional action recognition methods aim to recognize actions with complete observations/executions. However, it is often difficult to capture fully executed actions due to occlusions, interruptions, etc. Meanwhile, action prediction/recognition in advance based on partial observations is essential for preventing the situation from deteriorating. Besides, fast spotting human activities using partially observed data is a critical ingredient for retrieval systems. Inspired by the recent success of data binarization in efficient retrieval/recognition, we propose a novel approach, named Partial Reconstructive Binary Coding (PRBC), for action analysis based on limited frame glimpses during any period of the complete execution. Specifically, we learn discriminative compact binary codes for partial actions via a joint learning framework, which collaboratively tackles feature reconstruction as well as binary coding. We obtain the solution to PRBC based on a discrete alternating iteration algorithm. Extensive experiments on four realistic action datasets in terms of three tasks (i.e., partial action retrieval, recognition and prediction) clearly show the superiority of PRBC over the state-of-the-art methods, along with significantly reduced memory load and computational costs during the online test.

In this paper, we address the problem of cross-view image geo-localization. Specifically, we aim to estimate the GPS location of a query street view image by finding the matching images in a reference database of geo-tagged birds eye view images, or vice versa. To this end, we present a new framework for cross-view image geo-localization by taking advantage of the tremendous success of deep convolutional neural networks (CNNs) in image classification and object detection. First, we employ the Faster R-CNN to detect buildings in the query and reference images. Next, for each building in the query image, we retrieve the k nearest neighbors from the reference buildings using a Siamese network trained on both positive matching image pairs and negative pairs. To find the correct NN for each query building, we develop an efficient multiple nearest neighbors matching method based on dominant sets. We evaluate the proposed framework on a new dataset that consists of pairs of street view and birds eye view images. Experimental results show that the proposed method achieves better geo-localization accuracy than other approaches and is able to generalize to images at unseen locations.

Metric learning is an important issue in person re-identification, and Mahalanobis-distance based metric learning methods prevail in this field. All of these approaches can be considered as equivalently projecting all samples to a new metric space and calculating the Euclidean distance there. However, the performance of distinguishing similar samples from dissimilar ones via absolute distance is limited. In this paper, we suggest using relative distance instead. We adopt a bi-target perspective. The core idea is to construct a virtual opposite target for each original target. Then, the similarity between a sample and the others is judged by using both the original and opposite targets of the sample. In this way, we propose a bi-target metric method, named TAICHI distance. Considering simplicity and efficiency, we follow the KISSME metric in this paper. Extensive evaluations on challenging datasets confirm the effectiveness of the proposed method.

Deep learning has been demonstrated to achieve excellent results for image classification and object detection. However, the impact of deep learning on video analysis has been limited due to complexity of video data and lack of annotations. Previous convolutional neural networks (CNN) based video action detection approaches usually consist of two major steps: frame-level action proposal generation and association of proposals across frames. Also, most of these methods employ two-stream CNN framework to handle spatial and temporal feature separately. In this paper, we propose an end-to-end deep network called Tube Convolutional Neural Network (T-CNN) for action detection in videos. The proposed architecture is a unified deep network that is able to recognize and localize action based on 3D convolution features. A video is first divided into equal length clips and next for each clip a set of tube proposals are generated based on 3D Convolutional Network (ConvNet) features. Finally, the tube proposals of different clips are linked together employing network flow and spatio-temporal action detection is performed using these linked video proposals. Extensive experiments on several video datasets demonstrate the superior performance of T-CNN for classifying and localizing actions in both trimmed and untrimmed videos compared to state-of-the-arts.

Deep learning features for video action recognition are usually learned from RGB/gray images, image gradients, and optical flows. The single modality of the input data can describe one characteristic of the human action such as appearance structure or motion information. In this paper, we propose a high efficient gradient boundary convolutional network (ConvNet) to simultaneously learn spatio-temporal feature from the single modality data of gradient boundaries. The gradient boundaries represent both local spacial structure and motion information of action video. The gradient boundaries also have less background noise compared to RGB/gray images and image gradients. Extensive experiments are conducted on two popular and challenging action benchmarks, the UCF101 and the HMDB51 action datasets. The proposed deep gradient boundary feature achieves competitive performances on both benchmarks.

Human action classification, which is vital for content-based video retrieval and human-machine interaction, finds problem in distinguishing similar actions. Previous works typically detect spatial-temporal interest points (STIPs) from action sequences and then adopt bag-of-visual words (BoVW) model to describe actions as numerical statistics of STIPs. Despite the robustness of BoVW, this model ignores the spatial-temporal layout of STIPs, leading to misclassification among different types of actions with similar numerical statistics of STIPs. Motivated by this, a time-ordered feature is designed to describe the temporal distribution of STIPs, which contains complementary structural information to traditional BoVW model. Moreover, a temporal refinement method is used to eliminate intra-variations among time-ordered features caused by performers' habits. Then a time-ordered BoVW model is built to represent actions, which encodes both numerical statistics and temporal distribution of STIPs. Extensive experiments on three challenging datasets, i.e., KTH, Rochster and UT-Interaction, validate the effectiveness of our method in distinguishing similar actions.

This paper presents an effective local spatial-temporal descriptor for action recognition from skeleton sequences. The unique property of our descriptor is that it takes the spatial-temporal discrimination and action speed variations into account, intending to solve the problems of distinguishing similar actions and identifying actions with different speeds in one goal. The entire algorithm consists of two stages. First, a frame selection method is used to remove noisy skeletons for a given skeleton sequence. From the selected skeletons, skeleton joints are mapped to a high dimensional space, where each point refers to kinematics, time label and joint label of a skeleton joint. To encode relative relationships among joints, pairwise points from the space are then jointly mapped to a new space, where each point encodes the relative relationships of skeleton joints. Second, Fisher Vector (FV) is employed to encode all points from the new space as a compact feature representation. To cope with speed variations in actions, an energy-based temporal pyramid is applied to form a multi-temporal FV representation, which is fed into a kernel-based extreme learning machine classifier for recognition. Extensive experiments on benchmark datasets consistently show that our method outperforms state-of-the-art approaches for skeleton-based action recognition.

It remains a challenge to efficiently represent spatial-temporal data for 3D action recognition. To solve this problem, this paper presents a new skeleton-based action representation using data visualization and convolutional neural networks, which contains four main stages. First, skeletons from an action sequence are mapped as a set of five dimensional points, containing three dimensions of location, one dimension of time label and one dimension of joint label. Second, these points are encoded as a series of color images, by visualizing points as RGB pixels. Third, convolutional neural networks are adopted to extract deep features from color images. Finally, action class score is calculated by fusing selected deep features. Extensive experiments on three benchmark datasets show that our method achieves state-of-the-art results.

Pedestrian detection, as an important task in video surveillance and forensics applications, has been widely studied. However, its performance is unsatisfactory especially in the low resolution conditions. In realistic scenarios, the size of pedestrians in the images is often small, and detection can be challenging. To solve this problem, this paper proposes a novel resolution-score discriminative surface method to investigate the variation behaviors of detection scores under different pedestrian and non-pedestrian image resolutions. The discriminative surface consists of a series of positive and negative resolution-score lines, and each of them is a connected line to depict the variation relationship between pedestrian's detection scores under various image resolutions. On this basis, the resolution-score discriminative surface can classify a resolution-score line as a pedestrian or not according to whether it lies in the positive or the negative region. Experimental results on two public datasets and one campus surveillance dataset demonstrate the effectiveness of the proposed method.

This paper presents a novel deep convolutional feature fusion (ConvFF) approach for high-resolution scene classification, characterizing the well-known deep convolutional neural network (ConvNet) approach. The proposed ConvFF approach starts by generating an initial feature representation of the original scenes under exploration from two deep ConvNets pre-trained on two different large amount of labeled data. After the pre-training phase, we fine tune the two deep ConvNets consisting of mainly objects and scenes respectively in a supervised manner using the target training images. Then we propose to fuse the extracted two types of convolutional features provided by the last fully-connected (FC) layer, respectively. Finally, the fused convolutional features are fed as input to a SVM classifier for classification. The proposed method is evaluated by using two challenging high-resolution scene datasets. Experimental results show that the proposed method can effectively extract complementary features of the scenes and capture local spatial patterns, consistently outperforming several state-of-the-art methods.

The paper investigates representation-based classification for multispectral imagery. Due to the limited spectral dimension, the performance may be limited, and, in general, it is difficult to discriminate different classes using multispectral imagery. Nonlinear band generation method is proposed to use which can provide additional spectral information for multispectral classification. Two classifiers, sparse representation-based classification (SRC) and Nearest Regularized Subspace (NRS) are evaluated on the generated datasets. The results show our approach can outperform other nonlinear method such as the traditional kernel method in terms of classification accuracy and computational cost.

Representation-residual-based classifiers have attracted much attention in recent years in hyperspectral image (HSI) classification. How to obtain the optimal representa-tion coefficients for the classification task is the key problem of these methods. In this letter, spatial-aware collaborative representation (CR) is proposed for HSI classification. In order to make full use of the spatial-spectral information, we propose a closed-form solution, in which the spatial and spectral features are both utilized to induce the distance-weighted regularization terms. Different from traditional CR-based HSI classification algorithms, which model the spatial feature in a preprocessing or postprocessing stage, we directly incorporate the spatial information by adding a spatial regularization term to the representation objective function. The experimental results on three HSI data sets verify that our proposed approach outperforms the state-of-the-art classifiers.

Face image super-resolution has attracted much attention in recent years. Many algorithms have been proposed. Among them, sparse representation (SR)-based face image super-resolution approaches are able to achieve competitive performance. However, these SR-based approaches only perform well under the condition that the input is noiseless or has small noise. When the input is corrupted by large noise, the reconstruction weights (or coefficients) of the input low-resolution (LR) patches using SR-based approaches will be seriously unstable, thus leading to poor reconstruction results. To this end, in this paper, we propose a novel SR-based face image super-resolution approach that incorporates smooth priors to enforce similar training patches having similar sparse coding coefficients. Specifically, we introduce the fused least absolute shrinkage and selection operator-based smooth constraint and locality-based smooth constraint to the least squares representation-based patch representation in order to obtain stable reconstruction weights, especially when the noise level of the input LR image is high. Experiments are carried out on the benchmark FEI face database and CMU+MIT face database. Visual and quantitative comparisons show that the proposed face image super-resolution method yields superior reconstruction results when the input LR face image is contaminated by strong noise.

Human action recognition is an important yet challenging task. This paper presents a low-cost descriptor called 3D histograms of texture (3DHoTs) to extract discriminant features from a sequence of depth maps. 3DHoTs are derived from projecting depth frames onto three orthogonal Cartesian planes, i.e., the frontal, side, and top planes, and thus compactly characterize the salient information of a specific action, on which texture features are calculated to represent the action. Besides this fast feature descriptor, a new multi-class boosting classifier (MBC) is also proposed to efficiently exploit different kinds of features in a unified framework for action classification. Compared with the existing boosting frameworks, we add a new multi-class constraint into the objective function, which helps to maintain a better margin distribution by maximizing the mean of margin, whereas still minimizing the variance of margin. Experiments on the MSRAction3D, MSRGesture3D, MSRActivity3D, and UTD-MHAD data sets demonstrate that the proposed system combining 3DHoTs and MBC is superior to the state of the art.

The performance of traditional face recognition systems is sharply reduced when encountered with a low-resolution (LR) probe face image. To obtain much more detailed facial features, some face super-resolution (SR) methods have been proposed in the past decade. The basic idea of a face image SR is to generate a high-resolution (HR) face image from an LR one with the help of a set of training examples. It aims at transcending the limitations of optical imaging systems. In this paper, we regard face image SR as an image interpolation problem for domain-specific images. A missing intensity interpolation method based on smooth regression with a local structure prior (LSP), named SRLSP for short, is presented. In order to interpolate the missing intensities in a target HR image, we assume that face image patches at the same position share similar local structures, and use smooth regression to learn the relationship between LR pixels and missing HR pixels of one position patch. Performance comparison with the state-of-the-art SR algorithms on two public face databases and some real-world images shows the effectiveness of the proposed method for a face image SR in general. In addition, we conduct a face recognition experiment on the extended Yale-B face database based on the super-resolved HR faces. Experimental results clearly validate the advantages of our proposed SR method over the state-of-the-art SR methods in face recognition application.

The goal of learning-based image super resolution (SR) is to generate a plausible and visually pleasing high-resolution (HR) image from a given low-resolution (LR) input. The SR problem is severely underconstrained, and it has to rely on examples or some strong image priors to reconstruct the missing HR image details. This paper addresses the problem of learning the mapping functions (i.e., projection matrices) between the LR and HR images based on a dictionary of LR and HR examples. Encouraged by recent developments in image prior modeling, where the state-of-the-art algorithms are formed with nonlocal self-similarity and local geometry priors, we seek an SR algorithm of similar nature that will incorporate these two priors into the learning from LR space to HR space. The nonlocal self-similarity prior takes advantage of the redundancy of similar patches in natural images, while the local geometry prior of the data space can be used to regularize the modeling of the nonlinear relationship between LR and HR spaces. Based on the above two considerations, we first apply the local geometry prior to regularize the patch representation, and then utilize the nonlocal means filter to improve the super-resolved outcome. Experimental results verify the effectiveness of the proposed algorithm compared with the state-of-the-art SR methods.

The kernelized correlation filter (KCF) is one of the state-of-the-art object trackers. However, it does not reasonably model the distribution of correlation response during tracking process, which might cause the drifting problem, especially when targets undergo significant appearance changes due to occlusion, camera shaking, and/or deformation. In this paper, we propose an output constraint transfer (OCT) method that by modeling the distribution of correlation response in a Bayesian optimization framework is able to mitigate the drifting problem. OCT builds upon the reasonable assumption that the correlation response to the target image follows a Gaussian distribution, which we exploit to select training samples and reduce model uncertainty. OCT is rooted in a new theory which transfers data distribution to a constraint of the optimized variable, leading to an efficient framework to calculate correlation filters. Extensive experiments on a commonly used tracking benchmark show that the proposed method significantly improves KCF, and achieves better performance than other state-of-the-art trackers. To encourage further developments, the source code is made available.

In order to efficiently recognize actions from depth sequences, we propose a novel feature, called Global Ternary Image (GTI), which implicitly encodes both motion regions and motion directions between consecutive depth frames via recording the changes of depth pixels. In this study, each pixel in GTI indicates one of the three possible states, namely positive, negative and neutral, which represents increased, decreased and same depth values, respectively. Since GTI is sensitive to the subject's speed, we obtain energy-based GTI (E-GTI) by extracting GTI from pairwise depth frames with equal motion energy. To involve temporal information among depth frames, we extract E-GTI using multiple settings of motion energy. Here, the noise can be effectively suppressed by describing E-GTIs using the Radon Transform (RT). The 3D action representation is formed as a result of feeding the hierarchical combination of RTs to the Bag of Visual Words model (BoVW). From the extensive experiments on four benchmark datasets, namely MSRAction3D, DHA, MSRGesture3D and SKIG, it is evident that the hierarchical E-GTI outperforms the existing methods in 3D action recognition. We tested our proposed approach on extended MSRAction3D dataset to further investigate and verify its robustness against partial occlusions, noise and speed.

This paper presents a human action recognition approach by the simultaneous deployment of a second generation Kinect depth sensor and a wearable inertial sensor. Three data modalities consisting of depth images, skeleton joint positions, and inertial signals are fused by utilizing three collaborative representation classifiers. A database consisting of 10 actions performed by 6 subjects is put together to carry out two types of testing of the developed fusion approach: subject-generic and subject-specific. The overall recognition rates obtained from both types of testing indicate recognition improvements when fusing all the data modalities compared to the situations when data modalities are used individually.

Hyperspectral image (HSI) classification has attracted much attention and extensive research efforts over the past decade. Due to few labeled samples versus high dimensional features, it is a challenging problem in practice. Recently, combining the pixel spectral information and the spatial (neighborhood) information has been verified to be effective for HSI classification. In this paper, we introduce a novel method for HSI classification using set-to-set distance (SSD). Based on the assumption that neighbor pixels tend to belong to the same class with high probability, we model a test pixel and its neighbor pixels as a testing set (or a neighbor set) inspired by bilateral filtering. Meanwhile, the training pixels belong to the same class are modeled as a training set. Therefore, the classification is based on comparisons of sets distances. Experiments on a real HSI dataset show that our proposed method outperforms a number of existing state-of-the-art approaches.

In real world surveillance application, the captured faces are often low resolution (LR) and corrupted by mixed Gaussian-impulse noise during the acquisition and transmission processes. In this paper, we propose an effective patch-based face super-resolution method to reconstruct a high resolution (HR) face image given an LR observation that is corrupted by mixed Gaussian-impulse noise. To represent the corrupted image patches, a sparse regularization combined with an l\ data fitting term is proposed. In the proposed model, both the patch reconstruction term and the regularization term are in the l\ norm form. As a result, the model is called norms. In addition, since image pixels have nonnegative intensities, we further add a nonnegative constraint to the patch representation model. Experimental results demonstrate that the proposed norms based method can achieve superior face super-resolution performance over several state-of-the-art approaches based on the objective results in terms of P-SNR, as well as the visual perceptual quality.

This paper presents a novel extended multi-structure local binary pattern (EMSLBP) approach for high-resolution image classification, generalizing the well-known local binary pattern (LBP) approach. In the proposed EMSLBP approach, three-coupled descriptors with multi-structure sampling are proposed to extract complementary features (pixel value and radial difference) from local image patches. The anisotropic features derived from elliptical sampling are also rotation invariant by averaging the histograms over rotational angles and combined with the isotropic features extracted from circular sampling. Experimental results show that the proposed method can effectively capture local spatial pattern and local contrast, consistently outperforming several state-of-the-art classification algorithms.

Face super-resolution has attracted much attention in recent years. Many algorithms have been proposed. Among them, sparse representation based face super-resolution approaches are able to achieve competitive performance. However, these sparse representation based approaches only perform well under the condition that the input is noiseless or has small noise. When the input is corrupted by large noise, the reconstruction weights of the input LR patches using sparse representation based approaches will be seriously unstable, thus leading to poor reconstruction results. To this end, in this paper, we propose a novel sparse representation based face super-resolution approach that incorporates a smooth prior to enforce similar training patches having similar sparse coding coefficients. Specifically, we introduce the fused Lasso to the least squares representation of the input LR image in order to obtain a stable sparse representation, especially when the noise level of the input LR image is high. Experiments are carried out on the benchmark FEI face dataset. Visual and quantitative comparisons show that the proposed face super-resolution method achieves comparable performance to the state-of-the-art methods under noiseless condition, and yields superior super-resolution results when the input LR face image is contaminated by strong noise.

Human action recognition has a wide range of applications including biometrics, surveillance, and human computer interaction. The use of multimodal sensors for human action recognition is steadily increasing. However, there are limited publicly available datasets where depth camera and inertial sensor data are captured at the same time. This paper describes a freely available dataset, named UTD-MHAD, which consists of four temporally synchronized data modalities. These modalities include RGB videos, depth videos, skeleton positions, and inertial signals from a Kinect camera and a wearable inertial sensor for a comprehensive set of 27 human actions. Experimental results are provided to show how this database can be used to study fusion approaches that involve using both depth camera data and inertial sensor data. This public domain dataset is of benefit to multimodality research activities being conducted for human action recognition by various research groups.

This paper presents a computationally efficient method for action recognition from depth video sequences. It employs the so called depth motion maps (DMMs) from three projection views (front, side and top) to capture motion cues and uses local binary patterns (LBPs) to gain a compact feature representation. Two types of fusion consisting of feature-level fusion and decision-level fusion are considered. In the feature-level fusion, LBP features from three DMMs are merged before classification while in the decision-level fusion, a soft decision-fusion rule is used to combine the classification outcomes. The introduced method is evaluated on two standard datasets and is also compared with the existing methods. The results indicate that it outperforms the existing methods and is able to process depth video sequences in real-time.

It is of great interest in exploiting texture information for classification of hyperspectral imagery (HSI) at high spatial resolution. In this paper, a classification paradigm to exploit rich texture information of HSI is proposed. The proposed framework employs local binary patterns (LBPs) to extract local image features, such as edges, corners, and spots. Two levels of fusion (i.e., feature-level fusion and decision-level fusion) are applied to the extracted LBP features along with global Gabor features and original spectral features, where feature-level fusion involves concatenation of multiple features before the pattern classification process while decision-level fusion performs on probability outputs of each individual classification pipeline and soft-decision fusion rule is adopted to merge results from the classifier ensemble. Moreover, the efficient extreme learning machine with a very simple structure is employed as the classifier. Experimental results on several HSI data sets demonstrate that the proposed framework is superior to some traditional alternatives.

This paper presents a fusion approach for improving human action recognition based on two differing modality sensors consisting of a depth camera and an inertial body sensor. Computationally efficient action features are extracted from depth images provided by the depth camera and from accelerometer signals provided by the inertial body sensor. These features consist of depth motion maps and statistical signal attributes. For action recognition, both feature-level fusion and decision-level fusion are examined by using a collaborative representation classifier. In the feature-level fusion, features generated from the two differing modality sensors are merged before classification, while in the decision-level fusion, the Dempster-Shafer theory is used to combine the classification outcomes from two classifiers, each corresponding to one sensor. The introduced fusion framework is evaluated using the Berkeley multimodal human action database. The results indicate that because of the complementary aspect of the data from these sensors, the introduced fusion approaches lead to 2% to 23% recognition rate improvements depending on the action over the situations when each sensor is used individually.

Reconstruction of hyperspectral imagery from spectral random projections is considered. Specifically, multiple predictions drawn for a pixel vector of interest are made from spatially neighboring pixel vectors within an initial non-predicted reconstruction. A two-phase hypothesis-generation procedure based on partitioning and merging of spectral bands according to the correlation coefficients between bands is proposed to fine-tune the hypotheses. The resulting prediction is used to generate a residual in the projection domain. This residual being typically more compressible than the original pixel vector leads to improved reconstruction quality. To appropriately weight the hypothesis predictions, a distance-weighted Tikhonov regularization to an ill-posed least-squares optimization is proposed. Experimental results demonstrate that the proposed reconstruction significantly outperforms alternative strategies not employing multihypothesis prediction.

In this paper, we propose a novel algorithm for fast single image super-resolution based on self-example learning and sparse representation. We propose an efficient implementation based on the K-singular value decomposition (SVD) algorithm, where we replace the exact SVD computation with a much faster approximation, and we employ the straightforward orthogonal matching pursuit algorithm, which is more suitable for our proposed self-example-learning-based sparse reconstruction with far fewer signals. The patches used for dictionary learning are efficiently sampled from the low-resolution input image itself using our proposed sample mean square error strategy, without an external training set containing a large collection of high- resolution images. Moreover, the l0-optimization-based criterion, which is much faster than l1-optimization-based relaxation, is applied to both the dictionary learning and reconstruction phases. Compared with other super-resolution reconstruction methods, our low- dimensional dictionary is a more compact representation of patch pairs and it is capable of learning global and local information jointly, thereby reducing the computational cost substantially. Our algorithm can generate high-resolution images that have similar quality to other methods but with an increase in the computational efficiency greater than hundredfold.