Recently, text-to-image generation models have achieved remarkable advancements, particularly with diffusion models facilitating high-quality image synthesis from textual descriptions. However, these models often struggle with achieving precise control over pixel-level layouts, object appearances, and global styles when using text prompts alone. To mitigate this issue, previous works introduce conditional images as auxiliary inputs for image generation, enhancing control but typically necessitating specialized models tailored to different types of reference inputs. In this paper, we explore a new approach to unify controllable generation within a single framework. Specifically, we propose the unified image-instruction adapter (UNIC-Adapter) built on the Multi-Modal-Diffusion Transformer architecture, to enable flexible and controllable generation across diverse conditions without the need for multiple specialized models. Our UNIC-Adapter effectively extracts multi-modal instruction information by incorporating both conditional images and task instructions, injecting this information into the image generation process through a cross-attention mechanism enhanced by Rotary Position Embedding. Experimental results across a variety of tasks, including pixel-level spatial control, subject-driven image generation, and style-image-based image synthesis, demonstrate the effectiveness of our UNIC-Adapter in unified controllable image generation.

Diffusion-based models, widely used in text-to-image generation, have proven effective in 2D representation learning. Recently, this framework has been extended to 3D self-supervised learning by constructing a conditional point generator for enhancing 3D representations. However, its performance remains constrained by the 3D diffusion model, which is trained on the available 3D datasets with limited size. We hypothesize that the robust capabilities of text-to-image diffusion models, particularly Stable Diffusion (SD), which is trained on large-scale datasets, can help overcome these limitations. To investigate this hypothesis, we propose PointSD, a framework that leverages the SD model for 3D self-supervised learning. By replacing the SD model's text encoder with a 3D encoder, we train a point-to-image diffusion model that allows point clouds to guide the denoising of rendered noisy images. With the trained point-to-image diffusion model, we use noise-free images as the input and point clouds as the condition to extract SD features. Next, we train a 3D backbone by aligning its features with these SD features, thereby facilitating direct semantic learning. Comprehensive experiments on downstream point cloud tasks and ablation studies demonstrate that the SD model can enhance point cloud self-supervised learning. Code is publicly available at https://github.com/wdttt/PointSD.

Deep point cloud registration methods encounter challenges due to partial overlaps and are heavily reliant on labeled data. In this paper, we propose ALCReg, an active label correction method for partial point cloud registration learning. ALCReg utilises a multimodal approach to generate pseudo labels, mitigating the cold-start issue in active learning. To ensure the diversity and representativeness of selected samples, we propose an inlier ratio based query strategy for manual correction. Furthermore, an innovative self-correction mechanism based on consistency is introduced, allowing the model to refine pseudo labels autonomously and further improve model performance. Experimental results on the 3DMatch and 3DLoMatch datasets demonstrate that ALCReg achieves comparable performance with the fully-supervised registration methods, even with only 5% of labeled samples, making it the first active learning method tailored for partial point cloud registration. Code is available at https://github.com/Jiang0903/ALCReg.

This paper investigates indoor point cloud semantic segmentation under scene-level annotation, which is less explored compared to methods relying on sparse point-level labels. In the absence of precise point-level labels, current methods first generate point-level pseudo-labels, which are then used to train segmentation models. However, generating accurate pseudo-labels for each point solely based on scene-level annotations poses a considerable challenge, substantially affecting segmentation performance. Consequently, to enhance accuracy, this paper proposes a high-quality pseudo-label generation framework by exploring contemporary multi-modal information and region-point semantic consistency. Specifically, with a cross-modal feature guidance module, our method utilizes 2D-3D correspondences to align point cloud features with corresponding 2D image pixels, thereby assisting point cloud feature learning. To further alleviate the challenge presented by the scene-level annotation, we introduce a region-point semantic consistency module. It produces regional semantics through a region-voting strategy derived from point-level semantics, which are subsequently employed to guide the point-level semantic predictions. Leveraging the aforementioned modules, our method can rectify inaccurate point-level semantic predictions during training and obtain high-quality pseudo-labels. Significant improvements over previous works on ScanNet v2 and S3DIS datasets under scene-level annotation can demonstrate the effectiveness. Additionally, comprehensive ablation studies validate the contributions of our approach's individual components.

Weakly supervised LiDAR semantic segmentation has made significant strides with limited labeled data. However, most existing methods focus on the network training under weak supervision, while efficient annotation strategies remain largely unexplored. To tackle this gap, we implement LiDAR semantic segmentation using scatter image annotation, effectively integrating an efficient annotation strategy with network training. Specifically, we propose employing scatter images to annotate LiDAR point clouds, combining a pre-trained optical flow estimation network with a foundational image segmentation model to rapidly propagate manual annotations into dense labels for both images and point clouds. Moreover, we propose ScatterNet, a network that includes three pivotal strategies to reduce the performance gap caused by such annotations. First, it utilizes dense semantic labels as supervision for the image branch, alleviating the modality imbalance between point clouds and images. Second, an intermediate fusion branch is proposed to obtain multimodal texture and structural features. Finally, a perception consistency loss is introduced to determine which information needs to be fused and which needs to be discarded during the fusion process. Extensive experiments on the nuScenes and SemanticKITTI datasets demonstrate that our method requires less than 0.02% of the labeled points to achieve over 95% of the performance of fully-supervised methods. Notably, our labeled points are only 5% of those used in the most advanced weakly supervised methods.

Multi-view camera-based 3D object detection has become popular due to its low cost, but accurately inferring 3D geometry solely from camera data remains challenging and may lead to inferior performance. Although distilling precise 3D geometry knowledge from LiDAR data could help tackle this challenge, the benefits of LiDAR information could be greatly hindered by the significant modality gap between different sensory modalities. To address this issue, we propose a Simulated multi-modal Distillation (SimDistill) method by carefully crafting the model architecture and distillation strategy. Specifically, we devise multi-modal architectures for both teacher and student models, including a LiDAR-camera fusion-based teacher and a simulated fusion-based student. Owing to the ``identical'' architecture design, the student can mimic the teacher to generate multi-modal features with merely multi-view images as input, where a geometry compensation module is introduced to bridge the modality gap. Furthermore, we propose a comprehensive multi-modal distillation scheme that supports intra-modal, cross-modal, and multi-modal fusion distillation simultaneously in the Bird's-eye-view space. Incorporating them together, our SimDistill can learn better feature representations for 3D object detection while maintaining a cost-effective camera-only deployment. Extensive experiments validate the effectiveness and superiority of SimDistill over state-of-the-art methods, achieving an improvement of 4.8% mAP and 4.1% NDS over the baseline detector. The source code will be released at https://github.com/ViTAE-Transformer/SimDistill.

Rotation invariance is an important requirement for point shape analysis. To achieve this, current state-of-the-art methods attempt to construct the local rotation-invariant representation through learning or defining the local reference frame (LRF). Although efficient, these LRF-based methods suffer from perturbation of local geometric relations, resulting in suboptimal local rotation invariance. To alleviate this issue, we propose a Local-consistent Transformation (LocoTrans) learning strategy. Specifically, we first construct the local-consistent reference frame (LCRF) by considering the symmetry of the two axes in LRF. In comparison with previous LRFs, our LCRF is able to preserve local geometric relationships better through performing local-consistent transformation. However, as the consistency only exists in local regions, the relative pose information is still lost in the intermediate layers of the network. We mitigate such a relative pose issue by developing a relative pose recovery (RPR) module. RPR aims to restore the relative pose between adjacent transformed patches. Equipped with LCRF and RPR, our LocoTrans is capable of learning local-consistent transformation and preserving local geometry, which benefits rotation invariance learning. Competitive performance under arbitrary rotations on both shape classification and part segmentation tasks and ablations can demonstrate the effectiveness of our method. Code will be available publicly at https://github.com/wdttt/LocoTrans.

LiDAR semantic segmentation (LSS) is a critical task in autonomous driving and has achieved promising progress. However, prior LSS methods are conventionally investigated and evaluated on datasets within the same domain in clear weather. The robustness of LSS models in unseen scenes and all weather conditions is crucial for ensuring safety and reliability in real applications. To this end, we propose UniMix, a universal method that enhances the adaptability and generalizability of LSS models. UniMix first leverages physically valid adverse weather simulation to construct a Bridge Domain, which serves to bridge the domain gap between the clear weather scenes and the adverse weather scenes. Then, a Universal Mixing operator is defined regarding spatial, intensity, and semantic distributions to create the intermediate domain with mixed samples from given domains. Integrating the proposed two techniques into a teacher-student framework, UniMix efficiently mitigates the domain gap and enables LSS models to learn weather-robust and domain-invariant representations. We devote UniMix to two main setups: 1) unsupervised domain adaption, adapting the model from the clear weather source domain to the adverse weather target domain; 2) domain generalization, learning a model that generalizes well to unseen scenes in adverse weather. Extensive experiments validate the effectiveness of UniMix across different tasks and datasets, all achieving superior performance over state-of-the-art methods. The code will be released.

ControlNet excels at creating content that closely matches precise contours in user-provided masks. However, when these masks contain noise, as a frequent occurrence with non-expert users, the output would include unwanted artifacts. This paper first highlights the crucial role of controlling the impact of these inexplicit masks with diverse deterioration levels through in-depth analysis. Subsequently, to enhance controllability with inexplicit masks, an advanced Shape-aware ControlNet consisting of a deterioration estimator and a shape-prior modulation block is devised. The deterioration estimator assesses the deterioration factor of the provided masks. Then this factor is used in a modulation block to adaptively adjust the model's contour-following ability, which helps it dismiss the noise part in the inexplicit masks. Extensive experiments prove its effectiveness in encouraging ControlNet to interpret inaccurate spatial conditions robustly rather than blindly following the given contours, suitable for diverse kinds of conditions. We showcase application scenarios like modifying shape priors and composable shape-controllable generation. Codes are available at github.

Existing video-audio understanding models are trained and evaluated in an intra-domain setting, facing performance degeneration in real-world applications where multiple domains and distribution shifts naturally exist. The key to video-audio domain generalization (VADG) lies in alleviating spurious correlations over multi-modal features. To achieve this goal, we resort to causal theory and attribute such correlation to confounders affecting both video-audio features and labels. We propose a DeVADG framework that conducts uni-modal and cross-modal deconfounding through back-door adjustment. DeVADG performs cross-modal disentanglement and obtains fine-grained confounders at both class-level and domain-level using half-sibling regression and unpaired domain transformation, which essentially identifies domain-variant factors and class-shared factors that cause spurious correlations between features and false labels. To promote VADG research, we collect a VADG-Action dataset for video-audio action recognition with over 5,000 video clips across four domains (e.g., cartoon and game) and ten action classes (e.g., cooking and riding). We conduct extensive experiments, i.e., multi-source DG, single-source DG, and qualitative analysis, validating the rationality of our causal analysis and the effectiveness of the DeVADG framework.

Recently, Transformer-based methods, which predict polygon points or Bezier curve control points for localizing texts, are popular in scene text detection. However, these methods built upon detection transformer framework might achieve sub-optimal training efficiency and performance due to coarse positional query modeling. In addition, the point label form exploited in previous works implies the reading order of humans, which impedes the detection robustness from our observation. To address these challenges, this paper proposes a concise Dynamic Point Text DEtection TRansformer network, termed DPText-DETR. In detail, DPText-DETR directly leverages explicit point coordinates to generate position queries and dynamically updates them in a progressive way. Moreover, to improve the spatial inductive bias of non-local self-attention in Transformer, we present an Enhanced Factorized Self-Attention module which provides point queries within each instance with circular shape guidance. Furthermore, we design a simple yet effective positional label form to tackle the side effect of the previous form. To further evaluate the impact of different label forms on the detection robustness in real-world scenario, we establish an Inverse-Text test set containing 500 manually labeled images. Extensive experiments prove the high training efficiency, robustness, and state-of-the-art performance of our method on popular benchmarks. The code and the Inverse-Text test set are available at https://github.com/ymy-k/DPText-DETR.

End-to-end text spotting aims to integrate scene text detection and recognition into a unified framework. Dealing with the relationship between the two sub-tasks plays a pivotal role in designing effective spotters. Although Transformer-based methods eliminate the heuristic postprocessing, they still suffer from the synergy issue between the sub-tasks and low training efficiency. In this paper, we present DeepSolo, a simple DETR-like baseline that lets a single Decoder with Explicit Points Solo for text detection and recognition simultaneously. Technically, for each text instance, we represent the character sequence as ordered points and model them with learnable explicit point queries. After passing a single decoder, the point queries have encoded requisite text semantics and locations, thus can be further decoded to the center line, boundary, script, and confidence of text via very simple prediction heads in parallel. Besides, we also introduce a text-matching criterion to deliver more accurate supervisory signals, thus enabling more efficient training. Quantitative experiments on public benchmarks demonstrate that DeepSolo outperforms previous state-of-the-art methods and achieves better training efficiency. In addition, DeepSolo is also compatible with line annotations, which require much less annotation cost than polygons. The code is available at https://github.com/ViTAE-Transformer/DeepSolo.

Impressive performance on point cloud semantic segmentation has been achieved by fully-supervised methods with large amounts of labelled data. As it is labour-intensive to acquire large-scale point cloud data with point-wise labels, many attempts have been made to explore learning 3D point cloud segmentation with limited annotations. Active learning is one of the effective strategies to achieve this purpose but is still under-explored. The most recent methods of this kind measure the uncertainty of each pre-divided region for manual labelling but they suffer from redundant information and require additional efforts for region division. This paper aims at addressing this issue by developing a hierarchical point-based active learning strategy. Specifically, we measure the uncertainty for each point by a hierarchical minimum margin uncertainty module which considers the contextual information at multiple levels. Then, a feature-distance suppression strategy is designed to select important and representative points for manual labelling. Besides, to better exploit the unlabelled data, we build a semi-supervised segmentation framework based on our active strategy. Extensive experiments on the S3DIS and ScanNetV2 datasets demonstrate that the proposed framework achieves 96.5% and 100% performance of fully-supervised baseline with only 0.07% and 0.1% training data, respectively, outperforming the state-of-the-art weakly-supervised and active learning methods. The code will be available at https://github.com/SmiletoE/HPAL.

In recent years, cross-modal domain adaptation has been studied on the paired 2D image and 3D LiDAR data to ease the labeling costs for 3D LiDAR semantic segmentation (3DLSS) in the target domain. However, in such a setting the paired 2D and 3D data in the source domain are still collected with additional effort. Since the 2D-3D projections can enable the 3D model to learn semantic information from the 2D counterpart, we ask whether we could further remove the need of source 3D data and only rely on the source 2D images. To answer it, this paper studies a new 3DLSS setting where a 2D dataset (source) with semantic annotations and a paired but unannotated 2D image and 3D LiDAR data (target) are available1. To achieve 3DLSS in this scenario, we propose Cross-Modal and Cross-Domain Learning (CoMoDaL). Specifically, our CoMoDaL aims at modeling 1) inter-modal cross-domain distillation between the unpaired source 2D image and target 3D LiDAR data, and 2) the intra-domain cross-modal guidance between the target 2D image and 3D LiDAR data pair. In CoMoDaL, we propose to apply several constraints, such as point-to-pixel and prototype-to pixel alignments, to associate the semantics in different modalities and domains by constructing mixed samples in two modalities. The experimental results on several datasets show that in the proposed setting, the developed CoMoDaL can achieve segmentation without the supervision of labeled LiDAR data. Ablations are also conducted to provide more analysis. Code will be available publicly2.

Relying on large-scale training data with pixel-level labels, previous edge detection methods have achieved high performance. However, it is hard to manually label edges accurately, especially for large datasets, and thus the datasets inevitably contain noisy labels. This label-noise issue has been studied extensively for classification, while still remaining under-explored for edge detection. To address the label-noise issue for edge detection, this paper proposes to learn Pixel-level Noise Transitions to model the label-corruption process. To achieve it, we develop a novel Pixel-wise Shift Learning (PSL) module to estimate the transition from clean to noisy labels as a displacement field. Exploiting the estimated noise transitions, our model, named PNT-Edge, is able to fit the prediction to clean labels. In addition, a local edge density regularization term is devised to exploit local structure information for better transition learning. This term encourages learning large shifts for the edges with complex local structures. Experiments on SBD and Cityscapes demonstrate the effectiveness of our method in relieving the impact of label noise. Codes will be available at github.com/DREAMXFAR/PNT-Edge.

Transformers have been recently explored for 3D point cloud understanding with impressive progress achieved. A large number of points, over 0.1 million, make the global self-attention infeasible for point cloud data. Thus, most methods propose to apply the transformer in a local region, e.g., spherical or cubic window. However, it still contains a large number of Query-Key pairs, which requires high computational costs. In addition, previous methods usually learn the query, key, and value using a linear projection without modeling the local 3D geometric structure. In this paper, we attempt to reduce the costs and model the local geometry prior by developing a new transformer block, named ConDaFormer. Technically, ConDaFormer disassembles the cubic window into three orthogonal 2D planes, leading to fewer points when modeling the attention in a similar range. The disassembling operation is beneficial to enlarging the range of attention without increasing the computational complexity, but ignores some contexts. To provide a remedy, we develop a local structure enhancement strategy that introduces a depth-wise convolution before and after the attention. This scheme can also capture the local geometric information. Taking advantage of these designs, ConDaFormer captures both long-range contextual information and local priors. The effectiveness is demonstrated by experimental results on several 3D point cloud understanding benchmarks. Our code will be available.

Pseudo-labels are widely employed in weakly supervised 3D segmentation tasks where only sparse ground-truth labels are available for learning. Existing methods often rely on empirical label selection strategies, such as confidence thresholding, to generate beneficial pseudo-labels for model training. This approach may, however, hinder the comprehensive exploitation of unlabeled data points. We hypothesize that this selective usage arises from the noise in pseudo-labels generated on unlabeled data. The noise in pseudo-labels may result in significant discrepancies between pseudo-labels and model predictions, thus confusing and affecting the model training greatly. To address this issue, we propose a novel learning strategy to regularize the generated pseudo-labels and effectively narrow the gaps between pseudo-labels and model predictions. More specifically, our method introduces an Entropy Regularization loss and a Distribution Alignment loss for weakly supervised learning in 3D segmentation tasks, resulting in an ERDA learning strategy. Interestingly, by using KL distance to formulate the distribution alignment loss, it reduces to a deceptively simple cross-entropy-based loss which optimizes both the pseudo-label generation network and the 3D segmentation network simultaneously. Despite the simplicity, our method promisingly improves the performance. We validate the effectiveness through extensive experiments on various baselines and large-scale datasets. Results show that ERDA effectively enables the effective usage of all unlabeled data points for learning and achieves state-of-the-art performance under different settings. Remarkably, our method can outperform fully-supervised baselines using only 1\% of true annotations. Code and model will be made publicly available at https://github.com/LiyaoTang/ERDA.

In this study, we delve into the task of few-shot Generative Domain Adaptation (GDA), which involves transferring a pre-trained generator from one domain to a new domain using only a few reference images. Inspired by the way human brains acquire knowledge in new domains, we present an innovative generator structure called $\textbf{Domain Re-Modulation (DoRM)}$. DoRM not only meets the criteria of $\textit{high quality}$, $\textit{large synthesis diversity}$, and $\textit{cross-domain consistency}$, which were achieved by previous research in GDA, but also incorporates $\textit{memory}$ and $\textit{domain association}$, akin to how human brains operate. Specifically, DoRM freezes the source generator and introduces new mapping and affine modules (M\&A modules) to capture the attributes of the target domain during GDA. This process resembles the formation of new synapses in human brains. Consequently, a linearly combinable domain shift occurs in the style space. By incorporating multiple new M\&A modules, the generator gains the capability to perform high-fidelity multi-domain and hybrid-domain generation. Moreover, to maintain cross-domain consistency more effectively, we introduce a similarity-based structure loss. This loss aligns the auto-correlation map of the target image with its corresponding auto-correlation map of the source image during training. Through extensive experiments, we demonstrate the superior performance of our DoRM and similarity-based structure loss in few-shot GDA, both quantitatively and qualitatively. Code will be available at https://github.com/wuyi2020/DoRM.

Abstract Recent studies have shown promising results on joint learning of local feature detectors and descriptors. To address the lack of ground-truth keypoint supervision, previous methods mainly inject appropriate knowledge about keypoint attributes into the network to facilitate model learning. In this paper, inspired by traditional corner detectors, we develop an end-to-end deep network, named Deep Corner, which adds a local similarity-based keypoint measure into a plain convolutional network. Deep Corner enables finding reliable keypoints and thus benefits the learning of the distinctive descriptors. Moreover, to improve keypoint localization, we first study previous multi-level keypoint detection strategies and then develop a multi-level U-Net architecture, where the similarity of features at multiple levels can be exploited effectively. Finally, to improve the invariance of descriptors, we propose a feature self-transformation operation, which transforms the learned features adaptively according to the specific local information. The experimental results on several tasks and comprehensive ablation studies demonstrate the effectiveness of our method and the involved components.

Quantum computers are next-generation devices that hold promise to perform calculations beyond the reach of classical computers. A leading method towards achieving this goal is through quantum machine learning, especially quantum generative learning. Due to the intrinsic probabilistic nature of quantum mechanics, it is reasonable to postulate that quantum generative learning models (QGLMs) may surpass their classical counterparts. As such, QGLMs are receiving growing attention from the quantum physics and computer science communities, where various QGLMs that can be efficiently implemented on near-term quantum machines with potential computational advantages are proposed. In this paper, we review the current progress of QGLMs from the perspective of machine learning. Particularly, we interpret these QGLMs, covering quantum circuit Born machines, quantum generative adversarial networks, quantum Boltzmann machines, and quantum variational autoencoders, as the quantum extension of classical generative learning models. In this context, we explore their intrinsic relations and their fundamental differences. We further summarize the potential applications of QGLMs in both conventional machine learning tasks and quantum physics. Last, we discuss the challenges and further research directions for QGLMs.

Recently, blind image quality assessment (BIQA) models based on deep neural networks (DNNs) have achieved impressive performance on existing datasets. However, due to the intrinsic imbalance property of the training set, not all distortions or images are handled equally well. Online hard example mining (OHEM) is a promising way to alleviate this issue. Inspired by the recent finding that network pruning disproportionately hampers the model's memorization of a tractable subset, atypical, low-quality, long-tailed samples, that are hard-to-memorize during training and easily “forgotten” during pruning, we propose an effective “plug-and-play” OHEM pipeline, especially for generalizable deep BIQA. Specifically, we train two parallel weight-sharing branches simultaneously, where one is full model and other is a “self-competitor” generated from the full model online by network pruning. Then, we leverage the prediction disagreement between the full model and its pruned variant (i.e., the self-competitor) to expose easily “forgettable” samples, which are therefore regarded as the hard ones. We then enforce the prediction consistency between the full model and its pruned variant to implicitly put more focus on these hard samples, which benefits the full model to recover forgettable information introduced by pruning. Extensive experiments across multiple datasets and BIQA models demonstrate that the proposed OHEM can improve the model performance and generalizability as measured by correlation numbers and group maximum differentiation (gMAD) competition. Our code are available at:https://github.com/wangzhihua520/IQA_with_OHEM

In recent years, the security of AI systems has drawn increasing research attention, especially in the medical imaging realm. To develop a secure medical image analysis (MIA) system, it is a must to study possible backdoor attacks (BAs), which can embed hidden malicious behaviors into the system. However, designing a unified BA method that can be applied to various MIA systems is challenging due to the diversity of imaging modalities (e.g., X-Ray, CT, and MRI) and analysis tasks (e.g., classification, detection, and segmentation). Most existing BA methods are designed to attack natural image classification models, which apply spatial triggers to training images and inevitably corrupt the semantics of poisoned pixels, leading to the failures of attacking dense prediction models. To address this issue, we propose a novel Frequency-Injection based Backdoor Attack method (FIBA) that is capable of delivering attacks in various MIA tasks. Specifically, FIBA leverages a trigger function in the frequency domain that can inject the low-frequency information of a trigger image into the poisoned image by linearly combining the spectral amplitude of both images. Since it preserves the semantics of the poisoned image pixels, FIBA can perform attacks on both classification and dense prediction models. Experiments on three benchmarks in MIA (i.e., ISIC-2019 [4] for skin lesion classification, KiTS-19 [17] for kidney tumor segmentation, and EAD-2019 [1] for endoscopic artifact detection), validate the effectiveness of FIBA and its superiority over stateof-the-art methods in attacking MIA models and bypassing backdoor defense. Source code will be available at code.

Currently, single image inpainting has achieved promising results based on deep convolutional neural networks. However, inpainting on stereo images with missing regions has not been explored thoroughly, which is also a significant but different problem. One crucial requirement for stereo image inpainting is stereo consistency. To achieve it, we propose an Iterative Geometry-Aware Cross Guidance Network (IGGNet). The IGGNet contains two key ingredients, i.e., a Geometry-Aware Attention(GAA) module and an Iterative Cross Guidance(ICG) strategy. The GAA module relies on the epipolar geometry cues and learns the geometry-aware guidance from one view to another, which is beneficial to make the corresponding regions in two views consistent. However, learning guidance from co-existing missing regions is challenging. To address this issue, the ICG strategy is proposed, which can alternately narrow down the missing regions of the two views in an iterative manner. Experimental results demonstrate that our proposed network outperforms the latest stereo image inpainting model and state-of-the-art single image inpainting models.

Depth completion aims to recover a dense depth map from the sparse depth data and the corresponding single RGB image. The observed pixels provide the significant guidance for the recovery of the unobserved pixels' depth. However, due to the sparsity of the depth data, the standard convolution operation, exploited by most of existing methods, is not effective to model the observed contexts with depth values. To address this issue, we propose to adopt the graph propagation to capture the observed spatial contexts. Specifically, we first construct multiple graphs at different scales from observed pixels. Since the graph structure varies from sample to sample, we then apply the attention mechanism on the propagation, which encourages the network to model the contextual information adaptively. Furthermore, considering the mutli-modality of input data, we exploit the graph propagation on the two modalities respectively to extract multi-modal representations. Finally, we introduce the symmetric gated fusion strategy to exploit the extracted multi-modal features effectively. The proposed strategy preserves the original information for one modality and also absorbs complementary information from the other through learning the adaptive gating weights. Our model, named Adaptive Context-Aware Multi-Modal Network (ACMNet), achieves the state-of-the-art performance on two benchmarks, i.e., KITTI and NYU-v2, and at the same time has fewer parameters than latest models. Our code is available at: https://github.com/sshan-zhao/ACMNet.

As an interesting and important problem in computer vision, learning-based video saliency detection aims to discover the visually interesting regions in a video sequence. Capturing the information within frame and between frame at different aspects (such as spatial contexts, motion information, temporal consistency across frames, and multiscale representation) is important for this task. A key issue is how to jointly model all these factors within a unified data-driven scheme in an end-to-end fashion. In this article, we propose an end-to-end spatiotemporal deep video saliency detection approach, which captures the information on spatial contexts and motion characteristics. Furthermore, it encodes the temporal consistency information across the consecutive frames by implementing a convolutional long short-term memory (Conv-LSTM) model. In addition, the multiscale saliency properties for each frame are adaptively integrated for final saliency prediction in a collaborative feature-pyramid way. Finally, the proposed deep learning approach unifies all the aforementioned parts into an end-to-end joint deep learning scheme. Experimental results demonstrate the effectiveness of our approach in comparison with the state-of-the-art approaches.

Domain generalization aims to learn from multiple source domains a predictive model that can generalize to unseen target domains. One essential problem in domain generalization is to learn discriminative domain-invariant features. To arrive at this, some methods introduce a domain discriminator through adversarial learning to match the feature distributions in multiple source domains. However, adversarial training can only guarantee that the learned features have invariant marginal distributions, while the invariance of conditional distributions is more important for prediction in new domains. To ensure the conditional invariance of learned features, we propose an entropy regularization term that measures the dependency between the learned features and the class labels. Combined with the typical task-related loss, e.g., cross-entropy loss for classification, and adversarial loss for domain discrimination, our overall objective is guaranteed to learn conditional-invariant features across all source domains and thus can learn classifiers with better generalization capabilities. We demonstrate the effectiveness of our method through comparison with state-of-the-art methods on both simulated and real-world datasets. Code is available at: https://github.com/sshan-zhao/DGviaER.

Supervised depth estimation has achieved high accuracy due to the advanced deep network architectures. Since the groundtruth depth labels are hard to obtain, recent methods try to learn depth estimation networks in an unsupervised way by exploring unsupervised cues, which are effective but less reliable than true labels. An emerging way to resolve this dilemma is to transfer knowledge from synthetic images with ground truth depth via domain adaptation techniques. However, these approaches overlook specific geometric structure of the natural images in the target domain (i.e., real data), which is important for high-performing depth prediction. Motivated by the observation, we propose a geometry-aware symmetric domain adaptation framework (GASDA) to explore the labels in the synthetic data and epipolar geometry in the real data jointly. Moreover, by training two image style translators and depth estimators symmetrically in an end-to-end network, our model achieves better image style transfer and generates high-quality depth maps. The experimental results demonstrate the effectiveness of our proposed method and comparable performance against the state-of-the-art.

A key problem in co-saliency detection is how to effectively model the interactive relationship of a whole image group and the individual perspective of each image in a united data-driven manner. In this paper, we propose a group-wise deep co-saliency detection approach to address the co-saliency object discovery problem based on the fully convolutional network (FCN). The proposed approach captures the group-wise interaction information for group images by learning a semantics-aware image representation based on a convolutional neural network, which adaptively learns the group-wise features for co-saliency detection. Furthermore, the proposed approach discovers the collaborative and interactive relationships between group-wise feature representation and single image individual feature representation, and model this in a collaborative learning framework. Then, we set up a unified deep learning scheme to jointly optimize the process of group-wise feature representation learning and the collaborative learning, leading to more reliable and robust co-saliency detection results. Finally, we present a graph Laplacian regularized nonlinear regression model for saliency refinement. Experimental results demonstrate the effectiveness of our approach in comparison with the state-of-the-art approaches.

In this paper, we propose an end-to-end group-wise deep co-saliency detection approach to address the co-salient object discovery problem based on the fully convolutional network (FCN) with group input and group output. The proposed approach captures the group-wise interaction information for group images by learning a semantics-aware image representation based on a convolutional neural network, which adaptively learns the group-wise features for co-saliency detection. Furthermore, the proposed approach discovers the collaborative and interactive relationships between group-wise feature representation and single-image individual feature representation, and model this in a collaborative learning framework. Finally, we set up a unified end-to-end deep learning scheme to jointly optimize the process of group-wise feature representation learning and the collaborative learning, leading to more reliable and robust co-saliency detection results. Experimental results demonstrate the effectiveness of our approach in comparison with the state-of-the-art approaches.

As an important and challenging problem in computer vision, learning based optical flow estimation aims to discover the intrinsic correspondence structure between two adjacent video frames through statistical learning. Therefore, a key issue to solve in this area is how to effectively model the multi-scale correspondence structure properties in an adaptive end-to-end learning fashion. Motivated by this observation, we propose an end-to-end multi-scale correspondence structure learning (MSCSL) approach for optical flow estimation. In principle, the proposed MSCSL approach is capable of effectively capturing the multi-scale inter-image-correlation correspondence structures within a multi-level feature space from deep learning. Moreover, the proposed MSCSL approach builds a spatial Conv-GRU neural network model to adaptively model the intrinsic dependency relationships among these multi-scale correspondence structures. Finally, the above procedures for correspondence structure learning and multi-scale dependency modeling are implemented in a unified end-to-end deep learning framework. Experimental results on several benchmark datasets demonstrate the effectiveness of the proposed approach.