Surface reconstruction has been widely studied in computer vision and graphics. However, existing surface reconstruction works struggle to recover accurate scene geometry when the input views are extremely sparse. To address this issue, we propose MeshSplat, a generalizable sparse-view surface reconstruction framework via Gaussian Splatting. Our key idea is to leverage 2DGS as a bridge, which connects novel view synthesis to learned geometric priors and then transfers these priors to achieve surface reconstruction. Specifically, we incorporate a feed-forward network to predict per-view pixel-aligned 2DGS, which enables the network to synthesize novel view images and thus eliminates the need for direct 3D ground-truth supervision. To improve the accuracy of 2DGS position and orientation prediction, we propose a Weighted Chamfer Distance Loss to regularize the depth maps, especially in overlapping areas of input views, and also a normal prediction network to align the orientation of 2DGS with normal vectors predicted by a monocular normal estimator. Extensive experiments validate the effectiveness of our proposed improvement, demonstrating that our method achieves state-of-the-art performance in generalizable sparse-view mesh reconstruction tasks.

Typical detection-free methods for image-to-point cloud registration leverage transformer-based architectures to aggregate cross-modal features and establish correspondences. However, they often struggle under challenging conditions, where noise disrupts similarity computation and leads to incorrect correspondences. Moreover, without dedicated designs, it remains difficult to effectively select informative and correlated representations across modalities, thereby limiting the robustness and accuracy of registration. To address these challenges, we propose a novel cross-modal registration framework composed of two key modules: the Iterative Agents Selection (IAS) module and the Reliable Agents Interaction (RAI) module. IAS enhances structural feature awareness with phase maps and employs reinforcement learning principles to efficiently select reliable agents. RAI then leverages these selected agents to guide cross-modal interactions, effectively reducing mismatches and improving overall robustness. Extensive experiments on the RGB-D Scenes v2 and 7-Scenes benchmarks demonstrate that our method consistently achieves state-of-the-art performance.

Recent advances in point cloud analysis have increasingly leveraged large-scale unlabeled data through self-supervised representation learning. Autoregressive models based on next-token prediction have shown strong performance, but they usually model point clouds as linear sequences, ignoring their inherent spatial structure. To address this limitation, we propose PointChain, a novel autoregressive paradigm inspired by human perception mechanisms, designed to better align with the structural properties of point cloud. Specifically, we introduce structural chain encoding, which models the understanding process as a global-to-local structural chain inference, preserving spatial relationships throughout the prediction sequence. During pre-training, we design two auxiliary tasks: a next-scale prediction task that encourages cross-scale reasoning, and a scale-level contrastive learning task that promotes semantic consistency across scales. These components guide the model to learn more discriminative and generalizable point cloud representations. Experiments on multiple benchmarks, using both Transformer and Mamba backbones, validate the effectiveness of our approach. PointChain achieves state-of-the-art performance on several downstream tasks, including 93.75% accuracy on the hardest split of ScanObjectNN without voting strategy.

Single object tracking aims to localize target object with specific reference modalities (bounding box, natural language or both) in a sequence of specific video modalities (RGB, RGB+Depth, RGB+Thermal or RGB+Event.). Different reference modalities enable various human-machine interactions, and different video modalities are demanded in complex scenarios to enhance tracking robustness. Existing trackers are designed for single or several video modalities with single or several reference modalities, which leads to separate model designs and limits practical applications. Practically, a unified tracker is needed to handle various requirements. To the best of our knowledge, there is still no tracker that can perform tracking with these above reference modalities across these video modalities simultaneously. Thus, in this paper, we present a unified tracker, UniSOT, for different combinations of three reference modalities and four video modalities with uniform parameters. Extensive experimental results on 18 visual tracking, vision-language tracking and RGB+X tracking benchmarks demonstrate that UniSOT shows superior performance against modality-specific counterparts. Notably, UniSOT outperforms previous counterparts by over 3.0% AUC on TNL2K across all three reference modalities and outperforms Un-Track by over 2.0% main metric across all three RGB+X video modalities.

Image-to-point cloud registration methods typically follow a coarse-to-fine pipeline, extracting patch-level correspondences and refining them into dense pixel-to-point matches. However, in scenes with repetitive patterns, images often lack sufficient 3D structural cues and alignment with point clouds, leading to incorrect matches. Moreover, prior methods usually overlook structural consistency, limiting the full exploitation of correspondences. To address these issues, we propose two novel modules: the Local Geometry Enhancement (LGE) module and the Graph Distribution Consistency (GDC) module. LGE enhances both image and point cloud features with normal vectors, injecting geometric structure into image features to reduce mismatches. GDC constructs a graph from matched points to update features and explicitly constrain similarity distributions. Extensive experiments and ablations on two benchmarks, RGB-D Scenes v2 and 7-Scenes, demonstrate that our approach achieves state-of-the-art performance in image-to-point cloud registration.

Occlusion, a pervasive challenge in Multi-Object Tracking (MOT) within complex scenes, severely degrades tracking performance. Current methods still face numerous problems when handling occlusions. Motion prediction struggles to accommodate diverse motion patterns, resulting in failure during short-term occlusions. Concurrently, appearance features possess insufficient discriminative power under occluded conditions, which frequently leads to identity switches following long-term occlusion. To enhance the performance of MOT under such challenging conditions, we propose an innovative Occlusion-Aware Multi-Object Tracking via Joint Diffusion Motion Prediction and Appearance Purification (OAMOT). For short-term occlusions, a Diffusion Motion Recovery Model (DMRM) is developed, which integrates residual and noise diffusion branches to recover trajectories precisely under varied motion patterns. For long-term occlusions, the SAM Appearance Purification for ReID (SAPR) module is proposed; the module leverages the mask generation mechanism of SAM as foreground attention to enhance feature discriminability. Furthermore, a Lightweight Attention Predictor (LAP) is integrated into the ReID network to achieve SAM-quality attention during inference without significant computational overhead. Experimental results on the MOT17, MOT20, and DanceTrack datasets demonstrate that the proposed OAMOT method outperforms current state-of-the-art multi-object tracking techniques across multiple evaluation metrics. The code can be available on https://github.com/wangtuo111/OAMOT.

In recent years, multi-view 3D object detection has drawn a lot of interest in autonomous driving. The projection from 2D image point to 3D world space is the core step of multi-view 3D object detection. However, the uncertainty of depth estimation and camera parameters can make the projected 3D position inaccurate, leading to a performance drop. To this end, we propose an Uncertainty-aware Projection Network (UPN), including a depth uncertainty-guided 3D point generation module and a camera-aware 3D position encoder. For the depth uncertainty-guided 3D point generation module, we predict the depth as a distribution to generate depth uncertainty-aware 3D points. For the camera-aware 3D position encoder, we alleviate the uncertainty of camera parameters by employing adaptive receptive fields to refine the position embedding. Based on these two modules, the proposed UPN can effectively deal with the performance drop caused by inaccurate 2D-3D projection. To the best of our knowledge, this is the first work to model the 2D-3D projection uncertainty for multi-view 3D object detection. The proposed UPN achieves 56.9% mAP and 65.4 % NDS on the nuScenes test dataset, demonstrating the superiority of our method.

Existing image-text retrieval methods mainly rely on region and word features to measure cross-modal similarities. Thus, dense cross-modal semantic alignment which matches regions and words becomes crucial. However, this is non-trivial due to the heterogeneity gap and the cross-modal attention used to achieve this alignment is inefficient. Towards solving this problem, we propose a novel framework that goes beyond the previous one-tower and two-tower frameworks to learn cross-modal consensus efficiently. The proposed framework does not align regions and words directly like existing methods but uses semantic prototypes as a bridge to attend specific contents with the same semantics among different modalities through semantic decoders, through which cross-modal semantic alignment is naturally achieved. Furthermore, we design a novel plug-and-play self-correction method based on optimal transport to alleviate the drawbacks of incomplete pairwise labels in existing multimodal datasets. On top of various base backbones, we carry out extensive experiments on two benchmark datasets,i.e., Flickr30K and MS-COCO, demonstrating the effectiveness, superiority and generalization of our method.

Monocular depth estimation within the diffusion-denoising paradigm demonstrates impressive generalization ability but suffers from low inference speed. Recent methods adopt a single-step deterministic paradigm to improve inference efficiency while maintaining comparable performance. However, they overlook the gap between generative and discriminative features, leading to suboptimal results. In this work, we propose DepthMaster, a single-step diffusion model designed to adapt generative features for the discriminative depth estimation task. First, to mitigate overfitting to texture details introduced by generative features, we propose a Feature Alignment module, which incorporates high-quality semantic features to enhance the denoising network's representation capability. Second, to address the lack of fine-grained details in the single-step deterministic framework, we propose a Fourier Enhancement module to adaptively balance low-frequency structure and high-frequency details. We adopt a two-stage training strategy to fully leverage the potential of the two modules. In the first stage, we focus on learning the global scene structure with the Feature Alignment module, while in the second stage, we exploit the Fourier Enhancement module to improve the visual quality. Through these efforts, our model achieves state-of-the-art performance in terms of generalization and detail preservation, outperforming other diffusion-based methods across various datasets. Our project page can be found at https://indu1ge.github.io/DepthMaster_page.

Natural language tracking aims to locate the target of a video based on a language description. The rich contextual information inside the language and video sequence is essential to describe the target movements and appearance variations. However, existing natural language trackers design a fixed-length memory to store historical target information, which merely uses limited context information and necessitates manually designed modules, resulting in sub-optimal localization performance and numerous computational costs. Inspired by the success of the state space model, we propose a novel Context-adaptive Mamba Tracker (CMTrack). It enjoys several merits. First, we propose a novel context-aware state space model that enables language features to serve as hidden states to interact with relevant image features adaptively. Second, CMTrack transfers the hidden states frame-by-frame to continuously incorporate contextual target information into language features, enabling context-adaptive language cues. Third, the proposed context-adaptive language cues can effectively capture the long-range behavior of the target and guide the tracker in locating the target accurately without any extra design. Finally, CMTrack provides a neat pipeline for training and tracking with linear complexity. Experimental results demonstrate that CMTrack achieves new state-of-the-art performance.

3D single object tracking (3D SOT) remains a challenging task due to the sparsity of point clouds, appearance variations caused by occlusions, and the difficulty of modeling long-term temporal context. Although recent Transformer-based approaches leverage memory mechanisms to propagate temporal information, their quadratic complexity and reliance on discrete historical snapshots limit both efficiency and temporal coherence. To address these limitations, we propose SSMTrack, a novel 3D SOT framework built upon state space models (SSMs), which efficiently models long-term temporal dependencies through a continuously evolving hidden state with linear complexity. Specifically, we introduce a serialization and bidirectional scanning (SBS) strategy to enhance intra-frame feature interactions and design a Target-Aware Encoder (TAE) to extract target cues while maintaining stable temporal representations. Furthermore, we propose a Temporal Causal Shape Learning (TCSL) mechanism that preserves critical historical information while adaptively integrating current inputs, progressively enriching target feature representations over time. Extensive experiments on three benchmark datasets demonstrate that SSMTrack achieves state-of-the-art performance with strong temporal coherence and high efficiency. The code will be released upon publication.

Estimating the depth map of an image in the wild is a challenging visual task. Compared to relative depth estimation, metric depth estimation attracts more attention due to its practical physical significance and critical applications in real-life scenarios. However, existing depth estimation methods typically focus only on generalization of relative depth, neglecting the importance of metric depth generalization. To address this challenge, we propose a novel monocular depth estimation method called ScaleDepth. It decomposes metric depth into scene scale and relative depth and predicts them through a semantic-aware scale prediction (SASP) module and an adaptive relative depth estimation (ARDE) module, respectively. Our proposed approach has several merits. First, the SASP module can implicitly combine structural and semantic features of the images to predict precise scene scales. Second, the ARDE module can adaptively estimate the relative depth distribution of each image within a normalized depth space. Third, our method achieves metric depth estimation for both indoor and outdoor scenes in a unified framework, without the need to set the depth range or fine-tune the model. Extensive experiments demonstrate that our method achieves competitive performance in indoor, outdoor, unconfined, and unseen scenes. Project page: https://ruijiezhu94.github.io/ScaleDepth.

This paper addresses the critical challenge of domain adaptation for LiDAR-based semantic segmentation, particularly the significant density disparities that emerge when transferring models from synthetic to real-world environments. We present DA2-LiDAR, a novel density-adaptive domain adaptation framework that bridges domain gaps through the construction of intermediate domains with density-varying point distributions. Our approach employs a simple yet effective masking strategy that systematically reduces density discrepancies between domains while extracting more effective supervisory signals, as well as preserving critical semantic information. The framework consists of three key components: (1) a Density Adaptation Module that establishes a continuous spectrum of intermediate domains through dataset-agnostic masking operations; (2) a Contextual Consistency Module that enforces relational coherence across differently masked variants of the same scan at varying degrees, providing additional supervision signals, enhancing the model's ability to extract features; and (3) a Semantic Preservation Module that mitigates information loss in heavily masked scans by reconstructing domain-specific data distributions. Extensive experiments on synthetic-to-real and other benchmarks demonstrate that DA2-LiDAR consistently outperforms state-of-the-art methods, achieving significant improvements in cross-domain generalization without requiring dataset-specific prior knowledge or introducing computational overhead.

While existing unsupervised domain adaptation (UDA) methods greatly enhance target domain performance in semantic segmentation, they often neglect network calibration quality, resulting in misalignment between prediction confidence and actual accuracy-a significant risk in safety-critical applications. Our key insight emerges from observing that performance degrades substantially when soft pseudo-labels replace hard pseudo-labels in cross-domain scenarios due to poor calibration, despite the theoretical equivalence of perfectly calibrated soft pseudo-labels to hard pseudo-labels. Based on this finding, we propose DA-Cal, a dedicated cross-domain calibration framework that transforms target domain calibration into soft pseudo-label optimization. DA-Cal introduces a Meta Temperature Network to generate pixel-level calibration parameters and employs bi-level optimization to establish the relationship between soft pseudo-labels and UDA supervision, while utilizing complementary domain-mixing strategies to prevent overfitting and reduce domain discrepancies. Experiments demonstrate that DA-Cal seamlessly integrates with existing self-training frameworks across multiple UDA segmentation benchmarks, significantly improving target domain calibration while delivering performance gains without inference overhead. The code will be released.

Visual tracking aims to automatically estimate the state of a target object in a video sequence, which is challenging especially in dynamic scenarios. Thus, numerous methods are proposed to introduce temporal cues to enhance tracking robustness. However, conventional CNN and Transformer architectures exhibit inherent limitations in modeling long-range temporal dependencies in visual tracking, often necessitating either complex customized modules or substantial computational costs to integrate temporal cues. Inspired by the success of the state space model, we propose a novel temporal modeling paradigm for visual tracking, termed State-aware Mamba Tracker (SMTrack), providing a neat pipeline for training and tracking without needing customized modules or substantial computational costs to build long-range temporal dependencies. It enjoys several merits. First, we propose a novel selective state-aware space model with state-wise parameters to capture more diverse temporal cues for robust tracking. Second, SMTrack facilitates long-range temporal interactions with linear computational complexity during training. Third, SMTrack enables each frame to interact with previously tracked frames via hidden state propagation and updating, which releases computational costs of handling temporal cues during tracking. Extensive experimental results demonstrate that SMTrack achieves promising performance with low computational costs.

Adversarial examples are well known to pose a security risk, when attacking deep learning models. While, most of existing adversarial attacks are designed to attack a single deep learning-based task, such as image classification. In practical scenarios, it is more necessary to study adversarial examples transferring across different vision tasks. However, it is challenging to create cross-task adversarial examples that can destroy multiple vision tasks at once due to unavailable various task-specific models and loss functions for attackers. To deal with this problem, we propose a Dual Attention-Guided Method (DAGM) for crafting cross-task adversarial examples by designing a spatial attention module and a channel attention module to capture overlapping discriminative regions and features that contribute to various tasks. Then we craft cross-task adversarial examples via reducing the dispersion (i.e., standard deviation) of feature maps re-weighted by both attention modules, which can destroy the overlapping discriminative regions and features for various tasks. Furthermore, to present theoretical explanation, we systematically analyze our method, and rigorously prove that both attention modules can provide better effectiveness of our adversarial examples, compared with existing cross-task adversarial attacks. Extensive experiments on two datasets demonstrate that our method can significantly degrade the performance of various tasks, even online CV APIs, and consistently outperform state-of-the-art methods by a large margin.

Mitochondria segmentation from electron microscopy (EM) images plays a crucial role in biological and medical research. However, models trained on source domains often suffer from performance degradation when applied to target domains due to domain shift. Unsupervised domain adaptation (UDA) methods have been proposed to address this issue, but they often overlook the reliability of pseudo-labels and the effectiveness of supervision signals. In this paper, we propose R4MITO, a novel UDA framework for robust mitochondria segmentation. First, we introduce Reliable Prototype Pseudo-labels to mitigate the inconsistency of class-level features between across domains by leveraging source prototypes to model target prototypes. Second, we devise Correlation-wise Consistency Regularization to exploit inter-pixel correlations, aligning agent-level correlations under various perturbations. Third, we propose Rank-aware Relationship Consistency Regularization to fully utilize the rich information encoded in inter-agent relationships by imposing rank-aware constraints on agent-ranking probability distributions. Extensive experiments on multiple EM datasets demonstrate the superiority of our R4MITO over existing state-of-the-art UDA methods for mitochondria segmentation.

The method for image-to-point cloud registration typically determines the rigid transformation using a coarse-to-fine pipeline. However, directly and uniformly matching image patches with point cloud patches may lead to focusing on incorrect noise patches during matching while ignoring key ones. Moreover, due to the significant differences between image and point cloud modalities, it may be challenging to bridge the domain gap without specific improvements in design. To address the above issues, we innovatively propose the Uncertainty-aware Hierarchical Matching Module (UHMM) and the Adversarial Modal Alignment Module (AMAM). Within the UHMM, we model the uncertainty of critical information in image patches and facilitate multi-level fusion interactions between image and point cloud features. In the AMAM, we design an adversarial approach to reduce the domain gap between image and point cloud. Extensive experiments and ablation studies on RGB-D Scene V2 and 7-Scenes benchmarks demonstrate the superiority of our method, making it a state-of-the-art approach for image-to-point cloud registration tasks.

Transformers have demonstrated impressive results for 3D point cloud semantic segmentation. However, the quadratic complexity of transformer makes computation costs high, limiting the number of points that can be processed simultaneously and impeding the modeling of long-range dependencies between objects in a single scene. Drawing inspiration from the great potential of recent state space models (SSM) for long sequence modeling, we introduce Mamba, an SSM-based architecture, to the point cloud domain and propose Pamba, a novel architecture with strong global modeling capability under linear complexity. Specifically, to make the disorderness of point clouds fit in with the causal nature of Mamba, we propose a multi-path serialization strategy applicable to point clouds. Besides, we propose the ConvMamba block to compensate for the shortcomings of Mamba in modeling local geometries and in unidirectional modeling. Pamba obtains state-of-the-art results on several 3D point cloud segmentation tasks, including ScanNet v2, ScanNet200, S3DIS and nuScenes, while its effectiveness is validated by extensive experiments.

Few-Shot Video Object Segmentation (FSVOS) aims to achieve accurate segmentation of video sequences supported by limited annotated images. In this work, we analyze the deficiencies inherent in the use of object prototypes and pixel features as references in previous methods. Then we shed light on that part features, with the ability to adapt to appearance variations and resist noise, are advantageous as representative reference features for aligning support images and query videos. Therefore, we propose a Part Agent Learning Network (PALN) to leverage part features from two aspects. First, we elaborately employ Optimal Transport algorithm with equal partition constraint to make part agents capable of dividing support objects into diverse parts in an adaptive manner. Second, we design a dedicated cache mechanism to learn temporal part agents as lightweight historic target representation to exploit temporal consistency. With the aid of these learned part agents, our PALN can effectively achieve support-query alignment and temporal alignment for accurate segmentation of query videos. Extensive experimental results on two challenging benchmarks demonstrate that our method performs favorably against state-of-the-art FSVOS methods.

Vision-Language models (VLMs) have shown great potential in enhancing open-world visual concept comprehension. Recent researches focus on an optimum multimodal collaboration strategy that significantly advances CLIP-based few-shot tasks. However, existing prompt-based solutions suffer from unidirectional information flow and increased parameters since they explicitly condition the vision prompts on textual prompts across different transformer layers using non-shareable coupling functions. To address this issue, we propose a Dual-shared mechanism based on LoRA (DsRA) that addresses VLM adaptation in low-data regimes. The proposed DsRA enjoys several merits. First, we design an inter-modal shared coefficient that focuses on capturing visual and textual shared patterns, ensuring effective mutual synergy between image and text features. Second, an intra-modal shared matrix is proposed to achieve efficient parameter fine-tuning by combining the different coefficients to generate layer-wise adapters placed in encoder layers. Our extensive experiments demonstrate that DsRA improves the generalizability under few-shot classification, base-to-new generalization, and domain generalization settings. Our code will be released soon.

Relative pose estimation provides a promising way for achieving object-agnostic pose estimation. Despite the success of existing 3D correspondence-based methods, the reliance on explicit feature matching suffers from small overlaps in visible regions and unreliable feature estimation for invisible regions. Inspired by humans’ ability to assemble two object parts that have small or no overlapping regions by considering object structure, we propose a novel Structure-Aware Correspondence Learning method for Relative Pose Estimation, which consists of two key modules. First, a structure-aware keypoint extraction module is designed to locate a set of kepoints that can represent the structure of objects with different shapes and appearance, under the guidance of a keypoint based image reconstruction loss. Second, a structure-aware correspondence estimation module is designed to model the intra-image and inter-image relationships between keypoints to extract structure-aware features for correspondence estimation. By jointly leveraging these two modules, the proposed method can naturally estimate 3D-3D correspondences for unseen objects without explicit feature matching for precise relative pose estimation. Experimental results on the CO3D, Objaverse and LineMOD datasets demonstrate that the proposed method significantly outperforms prior methods, i.e., with 5.7° reduction in mean angular error on the CO3D dataset.

Video-Text Retrieval (VTR) is a core task in multi-modal understanding, drawing growing attention from both academia and industry in recent years. While numerous VTR methods have achieved success, most of them assume accurate visual-text correspondences during training, which is difficult to ensure in practice due to ubiquitous noise, known as noisy correspondences (NC). In this paper, we rethink how to mitigate the NC from the perspective of representative reference features (termed agents), and propose a novel relation-aware purified consistency (RPC) network to amend direct pairwise correlation, including representative agents construction and relation-aware ranking distribution alignment. The proposed RPC enjoys several merits. First, to learn the agents well without any correspondence supervision, we customize the agents construction according to the three characteristics of reliability, representativeness, and resilience. Second, the ranking distribution-based alignment process leverages the structural information inherent in inter-pair relationships, making it more robust compared to individual comparisons. Extensive experiments on five datasets under different settings demonstrate the efficacy and robustness of our method.

Cross-Domain Few-Shot Segmentation (CD-FSS) extends the generalization ability of Few-Shot Segmentation (FSS) beyond a single domain, enabling more practical applications. However, directly employing conventional FSS methods suffers from severe performance degradation in cross-domain settings, primarily due to feature sensitivity and support-to-query matching process sensitivity across domains. Existing methods for CD-FSS either focus on domain adaptation of features or delve into designing matching strategies for enhanced cross-domain robustness. Nonetheless, they overlook the fact that these two issues are interdependent and should be addressed jointly. In this work, we tackle these two issues within a unified framework by optimizing features in the frequency domain and enhancing the matching process in the spatial domain, working jointly to handle the deviations introduced by the domain gap. To this end, we propose a coherent Dual-Agent Optimization (DATO) framework, including a consistent mutual aggregation (CMA) and a correlation rectification strategy (CRS). In the consistent mutual aggregation module, we employ a set of agents to learn domain-invariant features across domains, and then use these features to enhance the original representations for feature adaptation. In the correlation rectification strategy, the agent-aggregated domain-invariant features serve as a bridge, transforming the support-to-query matching process into a referable feature space and reducing its domain sensitivity. Extensive experiments demonstrate the efficacy of our approach.

Image-to-point cloud registration aims to estimate the camera pose of a given image within a 3D scene point cloud. In this area, matching-based methods have achieved leading performance by first detecting the overlapping region, then matching point and pixel features learned by neural networks and finally using the PnP-RANSAC algorithm to estimate camera pose. However, achieving accurate image-to-point cloud registration remains challenging because the overlapping region detection is unreliable merely relying on point-wise classification, direct alignment of cross-modal data is difficult and indirect optimization objective leads to unstable registration results. To address these challenges, we propose a novel implicit correspondence learning method, including a Geometric Prior-guided overlapping region Detection Module (GPDM), an Implicit Correspondence Learning Module (ICLM), and a Pose Regression Module (PRM). The proposed method enjoys several merits. First, the proposed GPDM can precisely detect the overlapping region. Second, the ICLM can generate robust cross-modality correspondences. Third, the PRM can enable end-to-end optimization. Extensive experimental results on KITTI and nuScenes datasets demonstrate that the proposed model sets a new state-of-the-art performance in registration accuracy.

Category-level object pose estimation aims to determine the pose and size of arbitrary objects within given categories. Existing two-stage correspondence-based methods first establish correspondences between camera and object coordinates, and then acquire the object pose using a pose fitting algorithm. In this paper, we conduct a comprehensive analysis of this paradigm and introduce two crucial essentials: 1) shape-sensitive and pose-invariant feature extraction for accurate correspondence prediction, and 2) outlier correspondence removal for robust pose fitting. Based on these insights, we propose a simple yet effective correspondence-based method called SpotPose, which includes two stages. During the correspondence prediction stage, pose-invariant geometric structure of objects is thoroughly exploited to facilitate shape-sensitive holistic interaction among keypoint-wise features. During the pose fitting stage, outlier scores of correspondences are explicitly predicted to facilitate efficient identification and removal of outliers. Experimental results on CAMERA25, REAL275 and HouseCat6D benchmarks demonstrate that the proposed SpotPose outperforms state-of-the-art approaches by a large margin.

Detection-free methods typically follow a coarse-to-fine pipeline, extracting image and point cloud features for patch-level matching and refining dense pixel-to-point correspondences. However, differences in feature channel attention between images and point clouds may lead to degraded matching results, ultimately impairing registration accuracy. Furthermore, similar structures in the scene could lead to redundant correspondences in cross-modal matching. To address these issues, we propose Channel Adaptive Adjustment Module (CAA) and Global Optimal Selection Module (GOS). CAA enhances intra-modal features and suppresses cross-modal sensitivity, while GOS replaces local selection with global optimization. Experiments on RGB-D Scenes V2 and 7-Scenes demonstrate the superiority of our method, achieving state-of-the-art performance in image-to-point cloud registration.

The open vocabulary capability of 3D models is increasingly valued, as traditional methods with models trained with fixed categories fail to recognize unseen objects in complex dynamic 3D scenes. In this paper, we propose a simple yet effective approach, SAS, to integrate the open vocabulary capability of multiple 2D models and migrate it to 3D domain. Specifically, we first propose Model Alignment via Text to map different 2D models into the same embedding space using text as a bridge. Then, we propose Annotation-Free Model Capability Construction to explicitly quantify the 2D model's capability of recognizing different categories using diffusion models. Following this, point cloud features from different 2D models are fused with the guide of constructed model capabilities. Finally, the integrated 2D open vocabulary capability is transferred to 3D domain through feature distillation. SAS outperforms previous methods by a large margin across multiple datasets, including ScanNet v2, Matterport3D, and nuScenes, while its generalizability is further validated on downstream tasks, e.g., gaussian segmentation and instance segmentation.

Recently, Mamba-based methods have demonstrated impressive performance in point cloud representation learning by leveraging State Space Model (SSM) with the efficient context modeling ability and linear complexity. However, these methods still face two key issues that limit the potential of SSM: Destroying the adjacency of 3D points during SSM processing and failing to retain long-sequence memory as the input length increases in downstream tasks. To address these issues, we propose StruMamba3D, a novel paradigm for self-supervised point cloud representation learning. It enjoys several merits. First, we design spatial states and use them as proxies to preserve spatial dependencies among points. Second, we enhance the SSM with a state-wise update strategy and incorporate a lightweight convolution to facilitate interactions between spatial states for efficient structure modeling. Third, our method reduces the sensitivity of pre-trained Mamba-based models to varying input lengths by introducing a sequence length-adaptive strategy. Experimental results across four downstream tasks showcase the superior performance of our method. In addition, our method attains the SOTA 95.1% accuracy on ModelNet40 and 92.75% accuracy on the most challenging split of ScanObjectNN without voting strategy.

Category-level object pose estimation aims to determine the pose and size of novel objects in specific categories. Existing correspondence-based approaches typically adopt point-based representations to establish the correspondences between primitive observed points and normalized object coordinates. However, due to the inherent shape-dependence of canonical coordinates, these methods suffer from semantic incoherence across diverse object shapes. To resolve this issue, we innovatively leverage the sphere as a shared proxy shape of objects to learn shape-independent transformation via spherical representations. Based on this insight, we introduce a novel architecture called SpherePose, which yields precise correspondence prediction through three core designs. Firstly, We endow the point-wise feature extraction with SO(3)-invariance, which facilitates robust mapping between camera coordinate space and object coordinate space regardless of rotation transformation. Secondly, the spherical attention mechanism is designed to propagate and integrate features among spherical anchors from a comprehensive perspective, thus mitigating the interference of noise and incomplete point cloud. Lastly, a hyperbolic correspondence loss function is designed to distinguish subtle distinctions, which can promote the precision of correspondence prediction. Experimental results on CAMERA25, REAL275 and HouseCat6D benchmarks demonstrate the superior performance of our method, verifying the effectiveness of spherical representations and architectural innovations.

Semi-supervised semantic segmentation aims to learn from a limited amount of labeled data and a large volume of unlabeled data, which has witnessed impressive progress with the recent advancement of deep neural networks. However, existing methods tend to neglect the fact of class imbalance issues, leading to the Matthew effect, that is, the poorly calibrated model’s predictions can be biased towards the majority classes and away from minority classes with fewer samples. In this work, we analyze the Matthew effect present in previous methods that hinder model learning from a discriminative perspective. In light of this background, we integrate generative models into semi-supervised learning, taking advantage of their better class-imbalance tolerance. To this end, we propose DiffMatch to formulate the semi-supervised semantic segmentation task as a conditional discrete data generation problem to alleviate the Matthew effect of discriminative solutions from a generative perspective. Plus, to further reduce the risk of overfitting to the head classes and to increase coverage of the tail class distribution, we mathematically derive a debiased adjustment to adjust the conditional reverse probability towards unbiased predictions during each sampling step. Extensive experimental results across multiple benchmarks, especially in the most limited label scenarios with the most serious class imbalance issues, demonstrate that DiffMatch performs favorably against state-of-the-art methods.

DETR-based methods, which use multi-layer transformer decoders to refine object queries iteratively, have shown promising performance in 3D indoor object detection. However, the scene point features in the transformer decoder remain fixed, leading to minimal contributions from later decoder layers, thereby limiting performance improvement. Recently, State Space Models (SSM) have shown efficient context modeling ability with linear complexity through iterative interactions between system states and inputs. Inspired by SSMs, we propose a new 3D object DEtection paradigm with an interactive STate space model (DEST). In the interactive SSM, we design a novel state-dependent SSM parameterization method that enables system states to effectively serve as queries in 3D indoor detection tasks. In addition, we introduce four key designs tailored to the characteristics of point cloud and SSM: The serialization and bidirectional scanning strategies enable bidirectional feature interaction among scene points within the SSM. The inter-state attention mechanism models the relationships between state points, while the gated feed-forward network enhances inter-channel correlations. To the best of our knowledge, this is the first method to model queries as system states and scene points as system inputs, which can simultaneously update scene point features and query features with linear complexity. Extensive experiments on two challenging datasets demonstrate the effectiveness of our DEST-based method. Our method improves the GroupFree baseline in terms of $\text{AP}_{50}$ on ScanNet V2 (+5.3) and SUN RGB-D (+3.2) datasets. Based on the VDETR baseline, Our method sets a new state-of-the-art on the ScanNetV2 and SUN RGB-D datasets.

3D Instance segmentation (3DIS) on point clouds is a fundamental task in the field of 3D scene understanding. Existing fully-supervised networks have achieved promising results but remain heavily reliant on point-wise annotated data. Using the instance bounding boxes as annotations for weakly-supervised learning is a feasible way to solve the label-efficiency problem. In this paper, we propose POTNet, a novel training paradigm designed to generate point-wise pseudo-labels using only bounding box annotations, which effectively considering both local and global information. We leverage prototype learning method to extract local features from the non-overlapping regions indicated by the bounding boxes as instance prototypes. We employ an optimal transport algorithm to assign points in overlapping regions to their corresponding instances based on the similarity matrix between prototypes and these points. Our approach enables the generation of point-wise pseudo-labels that fully account for local and global correlations. We demonstrate the effectiveness of our method by achieving performance comparable to state-of-the-art approaches on multiple datasets, without requiring any additional supplementary data or retraining processes.

The critical challenge of semi-supervised semantic segmentation lies in how to fully exploit a large volume of unlabeled data to improve the model’s generalization performance for robust segmentation. Existing methods mainly rely on confidence-based scoring functions in the prediction space to filter pseudo labels, which suffer from the inherent trade-off between true and false positive rates. In this paper, we carefully design an agent construction strategy to build clean sets of correct (positive) and incorrect (negative) pseudo labels, and propose the Agent Score function (AgScore) to measure the consensus between candidate pixels and these sets. In this way, AgScore takes a step further to capture homogeneous patterns in the embedding space, conditioned on clean positive/negative agents stemming from the prediction space, without sacrificing the merits of confidence score, yielding better trad-off. We provide theoretical analysis to understand the mechanism of AgScore, and demonstrate its effectiveness by integrating it into three semi-supervised segmentation frameworks on Pascal VOC, Cityscapes, and COCO datasets, showing consistent improvements across all data partitions.

Unsupervised domain adaptation (UDA) for semantic segmentation aims to transfer knowledge from a labeled source domain to an unlabeled target domain. Despite the effectiveness of self-training techniques in UDA, they struggle to learn each class in a balanced manner due to inherent class imbalance and distribution shift in both data and label space between domains. To address this issue, we propose Balanced Learning for Domain Adaptation (BLDA), a novel approach to directly assess and alleviate class bias without requiring prior knowledge about the distribution shift. First, we identify over-predicted and under-predicted classes by analyzing the distribution of predicted logits. Subsequently, we introduce a post-hoc approach to align the logits distributions across different classes using shared anchor distributions. To further consider the network’s need to generate unbiased pseudo-labels during self-training, we estimate logits distributions online and incorporate logits correction terms into the loss function. Moreover, we leverage the resulting cumulative density as domain-shared structural knowledge to connect the source and target domains. Extensive experiments on two standard UDA semantic segmentation benchmarks demonstrate that BLDA consistently improves performance, especially for under-predicted classes, when integrated into various existing methods.

A series of pre-trained models have demonstrated promising results in point cloud understanding tasks and are widely applied to downstream tasks through fine-tuning. However, full fine-tuning leads to the forgetting of pretrained knowledge and substantial storage costs on edge devices. To address these issues, Parameter-Efficient Transfer Learning (PETL) methods have been proposed. According to our analysis, we find that existing 3D PETL methods cannot adequately align with semantic relationships of features required by downstream tasks, resulting in suboptimal performance. To ensure parameter efficiency while introducing rich semantic cues, we propose a novel fine-tuning paradigm for 3D pre-trained models. We utilize frozen 2D pre-trained models to provide vision semantic prompts and design a new Hybrid Attention Adapter to efficiently fuse 2D semantic cues into 3D representations with minimal trainable parameters(1.8M). Extensive experiments conducted on datasets including ScanObjectNN, ModelNet40, and ShapeNetPart demonstrate the effectiveness of our proposed paradigm. In particular, our method achieves 95.6% accuracy on ModelNet40 and attains 90.09% performance on the most challenging classification split ScanObjectNN(PB-T50-RS).

Point cloud understanding aims to acquire robust and general feature representations from unlabeled data. Masked point modeling-based methods have recently shown significant performance across various downstream tasks. These pre-training methods rely on random masking strategies to establish the perception of point clouds by restoring corrupted point cloud inputs, which leads to the failure of capturing reasonable semantic relationships by the self-supervised models. To address this issue, we propose Semantic Masked Autoencoder, which comprises two main components: a prototype-based component semantic modeling module and a component semantic-enhanced masking strategy. Specifically, in the component semantic modeling module, we design a component semantic guidance mechanism to direct a set of learnable prototypes in capturing the semantics of different components from objects. Leveraging these prototypes, we develop a component semantic-enhanced masking strategy that addresses the limitations of random masking in effectively covering complete component structures. Furthermore, we introduce a component semantic-enhanced prompt-tuning strategy, which further leverages these prototypes to improve the performance of pre-trained models in downstream tasks. Extensive experiments conducted on datasets such as ScanObjectNN, ModelNet40, and ShapeNetPart demonstrate the effectiveness of our proposed modules.

Leveraging Vision-Language Models (VLMs) like CLIP for various downstream tasks has emerged as a significant research trend. Recently, researchers have introduced Test-Time Adaptation (TTA) as a technique for models to learn online from unlabeled samples at test time, improving the generalization performance of VLMs to target domains. However, existing TTA methods either require expensive backpropagating gradient computations for each test sample or only extract knowledge from a limited number of historical test samples in the cache model, resulting in suboptimal adaptation performance. To address these limitations, we propose a Prototype Adaptive Fusion (PAF) framework, a novel TTA approach that makes full use of historical knowledge from test samples. Unlike traditional cache-based methods, which store only a few low-entropy samples per class, PAF introduces a prototype fusion mechanism that constructs class prototype representations through cumulatively merging features from qualified test samples. Furthermore, we propose an enhanced version, Easy-Hard PAF (EH-PAF), which adaptively applies a category-specific strategy based on CLIP prediction to improve performance. Extensive experiments across 15 diverse datasets demonstrate that our method consistently outperforms previous state-of-the-art approaches.

Domain adaptation for LiDAR semantic segmentation remains challenging due to the complex structural properties of point cloud data. While mix-based paradigms have shown promise, they often fail to fully leverage the rich structural priors inherent in 3D LiDAR point clouds. In this paper, we identify three critical yet underexploited structural priors: permutation invariance, local consistency, and geometric consistency. We introduce BeyondMix, a novel framework that harnesses the capabilities of State Space Models (specifically Mamba) to construct and exploit these structural priors while modeling long-range dependencies that transcend the limited receptive fields of conventional voxel-based approaches. By employing space-filling curves to impose sequential ordering on point cloud data and implementing strategic spatial partitioning schemes, BeyondMix effectively captures domain-invariant representations. Extensive experiments on challenging LiDAR semantic segmentation benchmarks demonstrate that our approach consistently outperforms existing state-of-the-art methods, establishing a new paradigm for unsupervised domain adaptation in 3D point cloud understanding.

Pseudo-label learning is widely used in semantic segmentation, particularly in label-scarce scenarios such as unsupervised domain adaptation (UDA) and semi-supervised learning (SSL). Despite its success, this paradigm can generate erroneous pseudo-labels, which are further amplified during training due to utilization of one-hot encoding. To address this issue, we propose ECOCSeg, a novel perspective for segmentation models that utilizes error-correcting output codes (ECOC) to create a fine-grained encoding for each class. ECOCSeg offers several advantages. First, an ECOC-based classifier is introduced, enabling model to disentangle classes into attributes and handle partial inaccurate bits, improving stability and generalization in pseudo-label learning. Second, a bit-level label denoising mechanism is developed to generate higher-quality pseudo-labels, providing adequate and robust supervision for unlabeled images. ECOCSeg can be easily integrated with existing methods and consistently demonstrates significant improvements on multiple UDA and SSL benchmarks across different segmentation architectures. Code is available at https://github.com/Woof6/ECOCSeg.

Semantic segmentation suffers from significant performance degradation when the trained network is applied to a different domain. To address this issue, unsupervised domain adaptation (UDA) has been extensively studied. Despite the effectiveness of selftraining techniques in UDA, they still overlook the explicit modeling of domain-shared feature extraction. In this paper, we propose DiDA, an unsupervised domain bridging approach for semantic segmentation. DiDA consists of two key modules: (1) Degradation-based Intermediate Domain Construction, which creates continuous intermediate domains through simple image degradation operations to encourage learning domain-invariant features as domain differences gradually diminish; (2) Semantic Shift Compensation, which leverages a diffusion encoder to disentangle and compensate for semantic shift information with degraded time-steps, preserving discriminative representations in the intermediate domains. As a plug-and-play solution, DiDA supports various degradation operations and seamlessly integrates with existing UDA methods. Extensive experiments on multiple domain adaptive semantic segmentation benchmarks demonstrate that DiDA consistently achieves significant performance improvements across all settings. Code is available at https://github.com/Woof6/DiDA.

Scene reconstruction from casually captured videos has wide real-world applications. Despite recent progress, existing methods relying on traditional cameras tend to fail in high-speed scenarios due to insufficient observations and inaccurate pose estimation. Event cameras, inspired by biological vision, record pixel-wise intensity changes asynchronously with high temporal resolution and low latency, providing valuable scene and motion information in blind inter-frame intervals. In this paper, we introduce the event cameras to aid scene construction from a casually captured video for the first time, and propose Event-Aided Free-Trajectory 3DGS, called EF-3DGS, which seamlessly integrates the advantages of event cameras into 3DGS through three key components. First, we leverage the Event Generation Model (EGM) to fuse events and frames, enabling continuous supervision between discrete frames. Second, we extract motion information through Contrast Maximization (CMax) of warped events, which calibrates camera poses and provides gradient-domain constraints for 3DGS. Third, to address the absence of color information in events, we combine photometric bundle adjustment (PBA) with a Fixed-GS training strategy that separates structure and color optimization, effectively ensuring color consistency across different views. We evaluate our method on the public Tanks and Temples benchmark and a newly collected real-world dataset, RealEv-DAVIS. Our method achieves up to 3dB higher PSNR and 40% lower Absolute Trajectory Error (ATE) compared to state-of-the-art methods under challenging high-speed scenarios.

Point cloud semantic segmentation is essential for understanding 3D scenes. Contemporary techniques often require extensive annotated training data, yet obtaining point-wise annotations for point clouds is time-consuming and laborious. Recent developments in weakly supervised methods seek to mitigate this problem by generating pseudo-labels using limited annotations. However, these pseudo-labels frequently suffer from either insufficient quantity or inferior quality. To overcome these hurdles, we introduce a Quantity-Quality Enhanced Self-training Network for Weakly Supervised Point Cloud Semantic Segmentation (Q2E). Specifically, an image-assisted pseudo-label generator is proposed to exploit 2D images to extend pseudo-labels for point clouds. Additionally, a hierarchical pseudo-label optimizer is developed to refine the quality of the pseudo-labels by hierarchically grouping them into broader categories. Extensive experiments on the ScanNet-v2, S3DIS, Semantic3D, and SemanticKITTI datasets demonstrate that Q2E outperforms state-of-the-art weakly supervised methods and rivals fully supervised approaches for point cloud semantic segmentation. Remarkably, as of the initial submission on February 2, 2024, our method ranked the first place in various settings of the ScanNet-v2 benchmark.

Geometry plays a significant role in monocular 3D object detection. It can be used to estimate object depth by using the perspective projection between object's physical size and 2D projection in the image plane, which can introduce mathematical priors into deep models. However, this projection process also introduces error amplification, where the error of the estimated height is amplified and reflected into the projected depth. It leads to unreliable depth inferences and also impairs training stability. To tackle this problem, we propose a novel Geometry Uncertainty Propagation Network (GUPNet++) by modeling geometry projection in a probabilistic manner. This ensures depth predictions are well-bounded and associated with a reasonable uncertainty. The significance of introducing such geometric uncertainty is two-fold: (1). It models the uncertainty propagation relationship of the geometry projection during training, improving the stability and efficiency of the end-to-end model learning. (2). It can be derived to a highly reliable confidence to indicate the quality of the 3D detection result, enabling more reliable detection inference. Experiments show that the proposed approach not only obtains (state-of-the-art) SOTA performance in image-based monocular 3D detection but also demonstrates superiority in efficacy with a simplified framework. The code and model will be released at https://github.com/SuperMHP/GUPNet_Plus.

Temporal action detection aims to predict temporal boundaries and category labels of actions in untrimmed videos. In the past years, many weakly supervised temporal action detection methods have been proposed to relieve the annotation cost of fully supervised methods. Due to the discrepancy between action localization and action classification, the two-branch structure is widely adopted by existing weakly supervised methods, where the classification branch is used to predict category-wise score and the localization branch is used to predict foreground score for each segment. Under the weakly supervised setting, the model training is mainly guided by the video-level or sparse segment-level annotations. As a result, the classification branch tends to focus on the most discriminative segments while ignore less discriminative ones so as to minimize the classification cost, and the localization branch may assign high foreground scores for some negative segments. This phenomenon can severely damage the action detection performance, because the foreground scores and classification scores are combined together in the testing stage for action detection. To deal with this problem, several methods have been proposed to encourage the consistency between the classification branch and localization branch. However, these methods only consider the video-level or segment-level consistency, without considering the relation among different segments to be consistent. In this paper, we propose a Cross-Task Relation-Aware Consistency (CRC) strategy for weakly supervised temporal action detection, including an intra-video consistency module and an inter-video consistency module. The intra-video consistency module can well guarantee the relationship among segments from the same video to be consistent, and the inter-video consistency module guarantees the relationship among segments from different videos to be consistent. These two modules are complementary to each other by combining both intra-video and inter-video consistency. Experimental results show that the proposed CRC strategy can consistently improve the performance of existing weakly supervised methods, including click-level supervised methods (e.g., LACP Lee et al., 2021), video-level supervised methods (e.g., DELU Chen et al., 2022) and unsupervised methods (e.g., BaS-Net Lee et al., 2020), verifying the generality and effectiveness of the proposed method.

Semi-supervised 3D object detection from point cloud aims to train a detector with a small number of labeled data and a large number of unlabeled data. Among existing methods, the pseudo-label based methods have achieved superior performance, and the core lies in how to select high-quality pseudo-labels with the designed quality evaluation criterion. Despite the success of these methods, they all consider the localization and classification quality estimation from a global perspective. For localization quality, they use a global score threshold to filter out low-quality pseudo-labels and assign equal importance to each side during training, ignoring the fact that sides with different localization quality should not be treat equally. Besides, a large number of pseudo-labels are discarded due to the high global threshold, which may also contain some correctly predicted sides that are helpful for model training. For the classification quality, they usually combine the objectness score and classification confidence score to filter out pseudo-labels. The main focus of them is designing effective classification confidence evaluation metrics, neglecting the importance of predicting better objectness score. In this paper, we propose SA3Det++, a side-aware quality estimation method for semi-supervised object detection, which consists of a probabilistic side localization strategy, a side-aware quality estimation strategy, and a soft pseudo-label selection strategy. Extensive results demonstrate that the proposed method consistently outperforms the baseline methods under different scenes and evaluation criterions.

Large-scale image-text pre-trained models have shown promising transferability to various downstream tasks. Video-text retrieval benefits from it by transferring pre-trained CLIP to video-text domain. Although these pre-trained models have shown impressive performance, full fine-tuning becomes prohibitively expensive as the size of these pre-trained models grows rapidly. To solve this, parameter-efficient tuning methods have been proposed, and prompt tuning is one of the most promising directions. However, existing prompt tuning methods do not have sufficient performance due to the lack of cross-modal interaction and prompt reliability assurance. To address these issues, we propose an effective and efficient Agent-based Control Prompt Tuning method (AbC-PT) for parameter-efficient video-text retrieval. The proposed AbC-PT enjoys several merits. Firstly, we design a parameter-efficient agent decoder with a carefully designed consistent attention mechanism to effectively capture video temporal information, mine contextual texts and perform cross-modal interaction between them. Secondly, we introduce two different sets of prompts, i.e., the vanilla prompt prepended to the input tokens and the concept prompt as the agent of the agent decoder. In addition, to ensure cross-modal semantic consistency of the concept prompt, we design a semantic consistency constraint loss. Thirdly, we devise a parameter-free prompt controller for adaptively calibrating each vanilla prompt based on its semantic in a data-driven way. Extensive experiments on five challenging benchmarks demonstrate that our method not only outperforms state-of-the-art parameter-efficient tuning methods, but even surpasses the full fine-tuning with 0.46% parameter overhead.

Monocular depth estimation aims to infer a dense depth map from a single image, which is a fundamental and prevalent task in computer vision. Many previous works have shown impressive depth estimation results through carefully designed network structures, but they usually ignore the planar information and therefore perform poorly in low-texture areas of indoor scenes. In this paper, we propose Plane2Depth, which adaptively utilizes plane information to improve depth prediction within a hierarchical framework. Specifically, in the proposed plane guided depth generator (PGDG), we design a set of plane queries as prototypes to softly model planes in the scene and predict per-pixel plane coefficients. Then the predicted plane coefficients can be converted into metric depth values with the pinhole camera model. In the proposed adaptive plane query aggregation (APGA) module, we introduce a novel feature interaction approach to improve the aggregation of multi-scale plane features in a top-down manner. Extensive experiments show that our method can achieve outstanding performance, especially in low-texture or repetitive areas. Furthermore, under the same backbone network, our method outperforms the state-of-the-art methods on the NYU-Depth-v2 dataset, achieves competitive results with state-of-the-art methods KITTI dataset and can be generalized to unseen scenes effectively.

Few-shot video object segmentation (FSVOS) aims to achieve accurate segmentation of novel objects in given video sequences, where the target objects are specified by limited annotated images as support. Most previous top-performing methods adopt the support-query semantic correlation learning paradigm or the intra-query temporal correlation learning paradigm. Nevertheless, they either fail to model temporal consistency across frames, resulting in inconsecutive segmentation, or lose diverse support object information, leading to incomplete segmentation. Therefore, we argue that it is more desirable to achieve both correlations in a collaborative manner. In this work, we delve into the issues present in the combination of few-shot image segmentation methods and video object segmentation methods and propose a dedicated Collaborative Correlation Network (CoCoNet) to address these problems, including a pixel correlation calibration module and a temporal correlation mining module. The proposed CoCoNet enjoys several merits. First, the pixel correlation calibration module aims to mitigate the noise issue in support-query correlation by integrating the affinity learning strategy and the prototype learning strategy. Specifically, we employ Optimal Transport to enrich pixel correlation with contextual information, thereby reducing intra-class differences between support and query. Second, the temporal correlation mining module is responsible for alleviating the issue of uncertainty in the initial frame and establishing reliable guidance for subsequent frames of the query video. With the collaboration of these two modules, our CoCoNet can effectively establish support-query and temporal correlation simultaneously and achieve accurate FSVOS. Extensive experimental results on two challenging benchmarks demonstrate that our method performs favorably against state-of-the-art FSVOS methods.

Open-vocabulary semantic segmentation (OVSS) aims to segment an image into regions of corresponding semantic vocabularies, without being limited to a predefined set of object categories. Existing works mainly utilize large-scale vision-language models (e.g., CLIP) to leverage their superior open-vocabulary classification abilities in a two-stage manner. However, their heavy reliance on the first-stage segmentation network leaves the full potential of CLIP untapped, creating an unresolved gap between the rich pre-training knowledge and the challenging per-pixel classification task. Although the recent one-stage paradigm has further leveraged pre-trained vision knowledge from CLIP, it fails to effectively utilize text information due to the inclusion of numerous unrelated semantics in the vocabulary list. How to avoid noise interference in text information and utilize language guidance remains a Gordian knot. In this paper, we propose a bi-directional bridge network (BBN) to bridge the gap between upstream pre-trained models and downstream segmentation tasks. It first purifies the noisy text embedding and then guides semantics-vision aggregation with the purified information in a purification-then-guidance manner, thereby facilitating effective semantic utilization. Specifically, we design an optimal purification modulator to purify noisy text information via the optimal transport algorithm, and a reliable guidance modulator to integrate proper textual information into vision embedding via the designed reliable attention in an adaptive manner. Extensive experimental results on five challenging benchmarks demonstrate that our BBN performs favorably against state-of-the-art open-vocabulary semantic segmentation methods.

Natural language tracking aims to locate the position of a target specified by a natural language description. Existing methods are trained on vision-language datasets with a small number of language descriptions, which may lead to limited semantic generalization. Moreover, they extract visual and language features separately, which limits visual-semantic capabilities. To overcome these limitations, we propose a novel semantic-aware tracking framework, SATrack, which integrates a semantic-aware attention module and a cross-modal aggregation module. The proposed SATrack enjoys several merits. First, the semantic-aware attention module utilizes language semantics as a bridge to build associations between visual features, enabling stronger visual-semantic capabilities. Second, the cross-modal aggregation module transfers the semantic knowledge of CLIP into the tracking framework for semantic generalization. Extensive experimental results demonstrate that SATrack outperforms previous state-of-the-art trackers on four natural language tracking benchmarks.

Cross-covariate gait recognition aims to analyze a pedestrian’s gait to extract an identity representation that is invariant across varying covariates. However, prevailing methods that have achieved good results on controlled in-the-lab datasets often perform poorly on realistic datasets. In this work, we find a significant cause is that the widely used pairwise metric learning paradigm cannot correctly handle the relationship between samples from different covariate conditions. Even worse, it may yield harmful signals that inadvertently mislead models to focus on covariate-related features, particularly when covariate distributions vary across subjects. To address this issue, we propose a Cross-Covariate Causal Intervention (GaitC3I) framework, a unified causality-inspired approach aimed at enhancing the robustness of gait recognition across diverse conditions. Specifically, our method consists of two parts: 1) an effective causal intervention metric learning paradigm based on backdoor adjustment, which strategically mitigates spurious correlations induced by covariates, thus ensuring a more invariant gait representation; and 2) an annotation-free selection strategy that progressively matches each positive sample with negative samples from similar covariate conditions at various granularities. We demonstrate the effectiveness of our GaitC3I through extensive evaluation on six popular gait datasets-Gait3D, GREW, OUMVLP, CASIA-B, CCPG, and CCGR-achieving substantial improvements. Our method not only outperforms existing state-of-the-art models but also provides a systematic solution to remove the spurious correlations in gait recognition.

3D single object tracking (3D SOT) in LiDAR point clouds plays a crucial role in autonomous driving. It remains a challenging problem due to the incompleteness and the sparsity of points caused by occlusion and limited sensor capabilities. Previous methods design various modules to propagate target perceptual cues to the current frame for target localization. However, perceptual cues may contain less information for occluded or distant objects, which brings great challenges to estimating the target state accurately. To address the above limitations, we propose a novel 3D SOT framework based on the adaptive conceptual prototypes named ACPTrack, which first learns the conceptual prototype from the prior knowledge of the category structure, and then associates weak perceptual cues with the learned conceptual prototypes to improve tracking performance. The proposed ACPTrack enjoys several merits. First, we propose a universal learning method of adaptive conceptual prototype, which can quickly adapt to target-specific structure with given perceptual cues. Second, we design two modules based on the conceptual prototype for structure completion and positioning refinement, which can exploit the rich structure information of the conceptual prototype to deal with sparse and incomplete targets for robust tracking. Third, our framework is generic and compatible with various 3D trackers and brings consistent performance gains. Extensive experiments validate that our method achieves competitive performance on three large-scale datasets.

Event cameras are bio-inspired sensors with diverse advantages, including high temporal resolution and minimal power consumption. Therefore, event cameras enjoy a wide range of applications in computer vision, among which event keypoint detection plays a vital role. However, repeatable event keypoint detection remains challenging because the lack of temporal interframe interaction leads to descriptors with limited temporal consistency, which restricts the ability to perceive keypoint motion. Besides, detectors learned at single scale features are not suitable for event keypoints with significant motion speed differences in high-speed scenarios. To deal with these problems, we propose a novel Spatio-Temporal Pyramid Keypoint Detection Network (STPNet) for event cameras via a temporally consistent descriptor learning (TCL) module and a spatially diverse detector learning (SDL) module. The proposed STPNet enjoys several merits. First, the TCL module generates temporally consistent descriptors for specific keypoint motion patterns. Second, the SDL module produces spatially diverse detectors for applications in high-speed motion scenarios. Extensive experimental results on three challenging benchmarks show that our method notably outperforms state-of-the-art event keypoint detection methods. Specifically, our STPNet can outperform the best event keypoint detection method by 0.21px in reprj. error on Event-Camera, 4% in IoU on N-Caltech101, 0.13px in reprj. error on HVGA ATIS Corner and 5.94% in matching accuracy on DSEC.

Weakly-supervised Temporal Action Localization (WTAL) aims to localize action instances with only video-level labels during training, where two primary issues are localization incompleteness and background interference. To relieve these two issues, recent methods adopt an attention mechanism to activate action instances and simultaneously suppress background ones, which have achieved remarkable progress. Nevertheless, we argue that these two issues have not been well resolved yet. On the one hand, the attention mechanism adopts fixed weights for different videos, which are incapable of handling the diversity of different videos, thus deficient in addressing the problem of localization incompleteness. On the other hand, previous methods only focus on learning the foreground attention and the attention weights usually suffer from ambiguity, resulting in difficulty of suppressing background interference. To deal with the above issues, in this paper we propose an Adaptive Prototype Learning (APL) method for WTAL, which includes two key designs: 1) an Adaptive Transformer Network (ATN) to explicitly model background and learn video-adaptive prototypes for each specific video; 2) an OT-based Collaborative (OTC) training strategy to guide the learning of prototypes and remove the ambiguity of the foreground-background separation by introducing an Optimal Transport (OT) algorithm into the collaborative training scheme between RGB and FLOW streams. These two key designs can work together to learn video-adaptive prototypes and solve the above two issues, achieving robust localization. Extensive experimental results on two standard benchmarks (THUMOS14 and ActivityNet) demonstrate that our proposed APL performs favorably against state-of-the-art methods.

Self-supervised point cloud representation learning aims to acquire robust and general feature representations from unlabeled data. Recently, masked point modeling-based methods have shown significant performance improvements for point cloud understanding, yet these methods rely on overlapping grouping strategies (k-nearest neighbor algorithm) resulting in early leakage of structural information of mask groups, and overlook the semantic modeling of object components resulting in parts with the same semantics having obvious feature differences due to position differences. In this work, we rethink grouping strategies and pretext tasks that are more suitable for self-supervised point cloud representation learning and propose a novel hierarchical masked representation learning method, including an optimal transport-based hierarchical grouping strategy, a prototype-based part modeling module, and a hierarchical attention encoder. The proposed method enjoys several merits. First, the proposed grouping strategy partitions the point cloud into non-overlapping groups, eliminating the early leakage of structural information in the masked groups. Second, the proposed prototype-based part modeling module dynamically models different object components, ensuring feature consistency on parts with the same semantics. Extensive experiments on four downstream tasks demonstrate that our method surpasses state-of-the-art 3D representation learning methods. Furthermore, Comprehensive ablation studies and visualizations demonstrate the effectiveness of the proposed modules.

Self-supervised monocular depth estimation holds significant importance in the fields of autonomous driving and robotics. However, existing methods are typically trained and evaluated on clear, sunny datasets, overlooking the impact of various adverse conditions commonly encountered in real-world applications, such as rainy weather, low visibility, and motion blur. As a result, they often struggle in challenging scenarios and produce artifacts. To address this issue, we propose ER-Depth, a novel two-stage self-supervised framework designed for robust depth estimation. In the first stage, we propose perturbation-invariant depth consistency regularization to propagate reliable supervision from standard to challenging scenes. In the second stage, we adopt the Mean Teacher paradigm for self-distillation and present a novel consistency-based pseudo-label filtering strategy to improve the quality of pseudo-labels. Extensive experiments demonstrate that our method exhibits exceptional robustness in challenging scenarios while maintaining high performance in standard scenes, significantly outperforming existing state-of-the-art methods on challenging KITTI-C, DrivingStereo, and NuScenes-Night benchmarks. Project page: https://ruijiezhu94.github.io/ERDepth_page .

Automatic mitochondrial segmentation enjoys great popularity with the development of deep learning. However, the coarse prediction raised by the presence of regular 3D grids in previous methods regardless of 3D CNN or the vision transformers suggest a possibly sub-optimal feature arrangement. To mitigate this limitation, we attempt to interpret the 3D EM image stacks as a set of interrelated 3D fragments for a better solution. However, it is non-trivial to model the 3D fragments without introducing excessive computational overhead. In this paper, we design a coherent fragment vision transformer (FragViT) combined with affinity learning to manipulate features on 3D fragments yet explore mutual relationships to model fragment-wise context, enjoying locality prior without sacrificing global reception. The proposed FragViT includes a fragment encoder and a hierarchical fragment aggregation module. The fragment encoder is equipped with affinity heads to transform the tokens into fragments with homogeneous semantics, and the multi-layer self-attention is used to explicitly learn inter-fragment relations with long-range dependencies. The hierarchical fragment aggregation module is responsible for hierarchically aggregating fragment-wise prediction back to the final voxel-wise prediction in a progressive manner. Extensive experimental results on the challenging MitoEM, Lucchi, and AC3/AC4 benchmarks demonstrate the effectiveness of the proposed method.

Single object tracking aims to locate the target object in a video sequence according to the state specified by different modal references, including the initial bounding box (BBOX), natural language (NL), or both (NL+BBOX). Due to the gap between different modalities, most existing trackers are designed for single or partial of these reference settings and overspecialize on the specific modality. Differently, we present a unified tracker called UVLTrack, which can simultaneously handle all three reference settings (BBOX, NL, NL+BBOX) with the same parameters. The proposed UVLTrack enjoys several merits. First, we design a modality-unified feature extractor for joint visual and language feature learning and propose a multi-modal contrastive loss to align the visual and language features into a unified semantic space. Second, a modality-adaptive box head is proposed, which makes full use of the target reference to mine ever-changing scenario features dynamically from video contexts and distinguish the target in a contrastive way, enabling robust performance in different reference settings. Extensive experimental results demonstrate that UVLTrack achieves promising performance on seven visual tracking datasets, three vision-language tracking datasets, and three visual grounding datasets. Codes and models will be open-sourced at https://github.com/OpenSpaceAI/UVLTrack.

Video semantic segmentation has achieved conspicuous achievements attributed to the development of deep learning, but suffers from labor-intensive annotated training data gathering. To alleviate the data-hunger issue, domain adaptation approaches are developed in the hope of adapting the model trained on the labeled synthetic videos to the real videos in the absence of annotations. By analyzing the dominant paradigm consistency regularization in the domain adaptation task, we find that the bottlenecks exist in previous methods from the perspective of pseudo-labels. To take full advantage of the information contained in the pseudo-labels and empower more effective supervision signals, we propose a coherent PAT network including a target domain focalizer and relation-aware temporal consistency. The proposed PAT network enjoys several merits. First, the target domain focalizer is responsible for paying attention to the target domain, and increasing the accessibility of pseudo-labels in consistency training. Second, the relation-aware temporal consistency aims at modeling the inter-class consistent relationship across frames to equip the model with effective supervision signals. Extensive experimental results on two challenging benchmarks demonstrate that our method performs favorably against state-of-the-art domain adaptive video semantic segmentation methods.

Cross-Domain Few-Shot Learning (CD-FSL) aims at recognizing samples in novel classes from unseen domains that are vastly different from training classes, with few labeled samples. However, the large domain gap between training and novel classes makes previous FSL methods perform poorly. To address this issue, we propose MetaPrompt, a Task-adaptive Prompted Transformer model for CD-FSL, by jointly exploiting prompt learning and the parameter generation framework. The proposed MetaPrompt enjoys several merits. First, a task-conditioned prompt generator is established upon attention mechanisms. It can flexibly produce a task-adaptive prompt with arbitrary length for unseen tasks, by selectively gathering task characteristics from the contextualized support embeddings. Second, the task-adaptive prompt is attached to Vision Transformer to facilitate fast task adaptation, steering the task-agnostic representation to incorporate task knowledge. To our best knowledge, this is the first work to exploit a prompt-based parameter generation mechanism for CD-FSL. Extensive experimental results on the Meta-Dataset benchmark demonstrate that our method achieves superior results against state-of-the-art methods.

Unsupervised point cloud shape correspondence aims to establish point-wise correspondences between source and target point clouds. Existing methods obtain correspon-dences directly by computing point-wise feature similar-ity between point clouds. However, non-rigid objects pos-sess strong deformability and unusual shapes, making it a longstanding challenge to directly establish correspon-dences between point clouds with unconventional shapes. To address this challenge, we propose an unsupervised Template-Assisted point cloud shape correspondence Net-work, termed TANet, including a template generation mod-ule and a template assistance module. The proposed TANet enjoys several merits. Firstly, the template generation mod-ule establishes a set of learnable templates with explicit structures. Secondly, we introduce a template assistance module that extensively leverages the generated templates to establish more accurate shape correspondences from multiple perspectives. Extensive experiments on four hu-man and animal datasets demonstrate that TANet achieves favorable performance against state-of-the-art methods.

Event cameras offer many advantages over traditional frame-based cameras, such as high dynamic range and low latency. Therefore, event cameras are widely applied in diverse computer vision applications, where event-based key-point detection is a fundamental task. However, achieving robust event-based keypoint detection remains challenging because the ground truth of event keypoints is difficult to obtain, descriptors extracted by CNN usually lack discrimi-native ability in the presence of intense noise, and fixed key-point detectors are limited in detecting varied keypoint patterns. To address these challenges, a novel event-based key-point detection method is proposed by learning dynamic de-tectors and contextual descriptors in a self-supervised manner (SD2Event), including a contextual feature descriptor learning (CFDL) module and a dynamic keypoint detector learning (DKDL) module. The proposed SD2Event enjoys several merits. First, the proposed CFDL module can model long-range contexts efficiently and effectively. Second, the DKDL module generates dynamic keypoint detectors, which can detect keypoints with diverse patterns across various event streams. Third, the proposed self-supervised signals can guide the model's adaptation to event data. Extensive experimental results on three challenging benchmarks show that our proposed method significantly outperforms state-of-the-art event-based keypoint detection methods.

Category-level 6D object pose estimation aims to estimate the rotation, translation and size of unseen instances within specific categories. In this area, dense correspondence-based methods have achieved leading performance. However, they do not explicitly consider the local and global geometric information of different instances, resulting in poor generalization ability to unseen instances with significant shape variations. To deal with this problem, we propose a novel Instance-Adaptive and Geometric-Aware Keypoint Learning method for category-level 6D object pose estimation (AG-Pose), which includes two key designs: (1) The first design is an Instance-Adaptive Keypoint Detection module, which can adaptively detect a set of sparse keypoints for various instances to represent their geometric structures. (2) The second design is a Geometric-Aware Feature Aggregation module, which can efficiently integrate the local and global geometric information into keypoint features. These two modules can work together to establish robust keypoint-level correspondences for unseen instances, thus enhancing the generalization ability of the model.Experimental results on CAMERA25 and REAL275 datasets show that the proposed AG-Pose outperforms state-of-the-art methods by a large margin without category-specific shape priors.

3D instance segmentation (3DIS) is a crucial task, but point-level annotations are tedious in fully supervised settings. Thus, using bounding boxes (bboxes) as annotations has shown great potential. The current mainstream approach is a two-step process, involving the generation of pseudo-labels from box annotations and the training of a 3DIS network with the pseudo-labels. However, due to the presence of intersections among bboxes, not every point has a determined instance label, especially in over-lapping areas. To generate higher quality pseudo-labels and achieve more precise weakly supervised 3DIS results, we propose the Box-Supervised Simulation-assisted Mean Teacher for 3D Instance Segmentation (BSNet), which devises a novel pseudo-labeler called Simulation-assisted Transformer. The labeler consists of two main components. The first is Simulation-assisted Mean Teacher, which introduces Mean Teacher for the first time in this task and constructs simulated samples to assist the labeler in acquiring prior knowledge about overlapping areas. To better model local-global structure, we also propose Local-Global Aware Attention as the decoder for teacher and student labelers. Extensive experiments conducted on the ScanNetV2 and S3DIS datasets verify the superiority of our designs. Code is available at https://github.com/peopleu/BSNet.

The key lie in semi-supervised semantic segmentation is how to fully exploit substantial unlabeled data to im-prove the model's generalization performance by resorting to constructing effective supervision signals. Most methods tend to directly apply contrastive learning to seek additional supervision to complement independent regular pixel-wise consistency regularization. However, these methods tend not to be preferred ascribed to their complicated designs, heavy memory footprints and susceptibility to confirmation bias. In this paper, we analyze the bottlenecks exist in con-trastive learning-based methods and offer a fresh perspective on inter-pixel correlations to construct more safe and effective supervision signals, which is in line with the nature of semantic segmentation. To this end, we develop a coherent RankMatch network, including the construction of representative agents to model inter-pixel correlation beyond regular individual pixel-wise consistency, and fur-ther unlock the potential of agents by modeling inter-agent relationships in pursuit of rank-aware correlation consis-tency. Extensive experimental results on multiple bench-marks, including mitochondria segmentation, demonstrate that RankMatch performs favorably against state-of-the-art methods. Particularly in the low-data regimes, RankMatch achieves significant improvements.

Open-vocabulary semantic segmentation (OVS) aims to segment images of arbitrary categories specified by class labels or captions. However, most previous best-performing methods, whether pixel grouping methods or region recognition methods, suffer from false matches between image features and category labels. We attribute this to the natural gap between the textual features and visual features. In this work, we rethink how to mitigate false matches from the perspective of image-to-image matching and propose a novel relation-aware intra-modal matching (RIM) framework for OVS based on visual foundation models. RIM achieves robust region classification by firstly constructing diverse image-modal reference features and then matching them with region features based on relation-aware ranking distribution. The proposed RIM enjoys several merits. First, the intra-modal reference features are better aligned, circumventing potential ambiguities that may arise in cross-modal matching. Second, the ranking-based matching process harnesses the structure information implicit in the inter-class relationships, making it more robust than comparing individually. Extensive experiments on three benchmarks demonstrate that RIM outperforms previous state-of-the-art methods by large margins, obtaining a lead of more than 10% in mIoU on PASCAL VOC benchmark.

Referring Remote Sensing Image Segmentation (RRSIS) is a challenging task that aims to identify specific regions in aerial images that are relevant to given textual conditions. Existing methods tend to adopt the paradigm of implicit optimization, utilizing a framework consisting of early cross-modal feature fusion and a fixed convolutional kernel-based predictor, neglecting the inherent inter-domain gap and conducting class-agnostic predictions. In this paper, we rethink the issues with the implicit optimization paradigm and address the RRSIS task from a dual-alignment perspective. Specifically, we prepend the dedicated Dual Alignment Network (DANet), including an explicit alignment strategy and a reliable agent alignment module. The explicit alignment strategy effectively reduces domain discrepancies by narrowing the inter-domain affinity distribution. Meanwhile, the reliable agent alignment module aims to enhance the predictor's multi-modality awareness and alleviate the impact of deceptive noise interference. Extensive experiments on two remote sensing datasets demonstrate the effectiveness of our proposed DANet in achieving superior segmentation performance without introducing additional learnable parameters compared to state-of-the-art methods.

Dynamic scene reconstruction is a long-term challenge in the field of 3D vision. Recently, the emergence of 3D Gaussian Splatting has provided new insights into this problem. Although subsequent efforts rapidly extend static 3D Gaussian to dynamic scenes, they often lack explicit constraints on object motion, leading to optimization difficulties and performance degradation. To address the above issues, we propose a novel deformable 3D Gaussian splatting framework called MotionGS, which explores explicit motion priors to guide the deformation of 3D Gaussians. Specifically, we first introduce an optical flow decoupling module that decouples optical flow into camera flow and motion flow, corresponding to camera movement and object motion respectively. Then the motion flow can effectively constrain the deformation of 3D Gaussians, thus simulating the motion of dynamic objects. Additionally, a camera pose refinement module is proposed to alternately optimize 3D Gaussians and camera poses, mitigating the impact of inaccurate camera poses. Extensive experiments in the monocular dynamic scenes validate that MotionGS surpasses state-of-the-art methods and exhibits significant superiority in both qualitative and quantitative results. Project page: https://ruijiezhu94.github.io/MotionGS_page.

Dynamic scenes rendering is an intriguing yet challenging problem. Although current methods based on NeRF have achieved satisfactory performance, they still can not reach real-time levels. Recently, 3D Gaussian Splatting (3DGS) has garnered researchers' attention due to their outstanding rendering quality and real-time speed. Therefore, a new paradigm has been proposed: defining a canonical 3D gaussians and deforming it to individual frames in deformable fields. However, since the coordinates of canonical 3D gaussians are filled with noise, which can transfer noise into the deformable fields, and there is currently no method that adequately considers the aggregation of 4D information. Therefore, we propose Denoised Deformable Network with Temporal-Spatial Aggregation for Dynamic Scene Rendering (DN-4DGS). Specifically, a Noise Suppression Strategy is introduced to change the distribution of the coordinates of the canonical 3D gaussians and suppress noise. Additionally, a Decoupled Temporal-Spatial Aggregation Module is designed to aggregate information from adjacent points and frames. Extensive experiments on various real-world datasets demonstrate that our method achieves state-of-the-art rendering quality under a real-time level. Code is available at https://github.com/peoplelu/DN-4DGS.

Video-Text Retrieval is a fundamental task in multi-modal understanding and has attracted increasing attention from both academia and industry communities in recent years. Generally, video inherently contains multi-grained semantic and each video corresponds to several different texts, which is challenging. Previous best-performing methods adopt video-sentence, phrase-phrase, and frame-word interactions simultaneously. Different from word/frame features that can be obtained directly, phrase features need to be adaptively aggregated from correlative word/frame features, which makes it very demanding. However, existing method utilizes simple intra-modal self-attention to generate phrase features without considering the following three aspects: cross-modality semantic correlation, phrase generation noise and diversity. In this paper, we propose a novel Reliable Phrase Mining model (RPM) to construct reliable phrase features and conduct hierarchical cross-modal interactions for video-text retrieval. The proposed RPM model enjoys several merits. Firstly, to guarantee the semantic consistency between video phrases and text phrases, we propose a set of modality-shared prototypes as the joint query to aggregate the semantically related frame/word features into adaptive-grained phrase features. Secondly, to deal with the phrase generation noise, the proposed denoised decoder module is responsible for obtaining more reliable similarity between prototypes and frame/word features. Specifically, not only the correlation between frame/word features and prototypes, but also the correlation among prototypes, should be taken into account when calculating the similarity. Furthermore, to encourage different prototypes to focus on different semantic information, we design a prototype contrastive loss whose core idea is enabling phrases produced by the same prototype to be more similar than those produced by different prototypes. Extensive experiment results demonstrate that the proposed method performs favorably on three benchmark datasets, including MSR-VTT, MSVD, and LSMDC.

Pre-trained Vision-Language Models (VLMs), like CLIP, exhibit strong generalization ability to downstream tasks but struggle in few-shot scenarios. Existing prompting techniques primarily focus on global text and image representations, yet overlooking multi-modal attribute characteristics. This limitation hinders the model’s ability to perceive fine-grained visual details and restricts its generalization ability to a broader range of unseen classes. To address this issue, we propose a Multi-modal Attribute Prompting method (MAP) by jointly exploring textual attribute prompting, visual attribute prompting, and attribute-level alignment. The proposed MAP enjoys several merits. First, we introduce learnable visual attribute prompts enhanced by textual attribute semantics to adaptively capture visual attributes for images from unknown categories, boosting fine-grained visual perception capabilities for CLIP. Second, the proposed attribute-level alignment complements the global alignment to enhance the robustness of cross-modal alignment for open-vocabulary objects. To our knowledge, this is the first work to establish cross-modal attribute-level alignment for CLIP-based few-shot adaptation. Extensive experimental results on 11 datasets demonstrate that our method performs favorably against state-of-the-art approaches.

Image-text matching aims to bridge vision and language areas, which is a crucial task in multi-modal intelligence. The core idea is to learn features of each modality and aggregate learned features as holistic representations to measure image-text relevance. Most existing methods involve cross-modal interaction during feature learning by modeling fine-grained relationships between two modalities for better results. However, these methods may obtain wrong attention scores when directly computing similarities between regions and words. Besides, current methods mainly rely on simple pooling operations for feature aggregation, which introduces interference from redundant information, resulting in inaccurate matching results. To alleviate these issues, we propose a novel reference-aware adaptive network for image-text matching by jointly using a reference attention module for feature learning and an adaptive aggregation module for feature aggregation. The proposed model enjoys several merits. First, the designed reference attention module effectively reduces wrong attention scores by introducing a set of references during cross-modal interaction. Second, the proposed adaptive aggregation module highlights useful information adaptively while suppressing redundant information during aggregation. Extensive experiments on two standard benchmarks demonstrate that our method performs favorably against state-of-the-art methods.

Few-Shot Learning (FSL) leverages prior knowledge and generalization strategies to quickly adapt to new tasks or recognize new objects with minimal input. Recently, CLIP-based methods, aided by contrastive language-image pre-training, have demonstrated impressive few-shot performance. However, these methods solely employ fixed-length uni-modal prompts at the initial encoder layer, neglecting the multi-level adaptation and cross-modal interaction for the intermediate features. To address this issue, we propose Hierarchy-Aware Interactive Prompt Learning (HIPL), by jointly exploring hierarchical prompt learning and cross-modal prompt interaction for CLIP-based FSC. The proposed HIPL enjoys several merits. First, we design a hierarchical prompt aggregation module to progressively generate higher-level prompts via the attention mechanisms, equipping the CLIP with hierarchical adaptation capability. Second, a cross-modal prompt interaction module is proposed to facilitate deep interaction between stage-wise prompts, ensuring mutual synergy between vision and textual features. To the best of our knowledge, this is the first work to learn multi-level prompts by progressive aggregation. Our extensive experiments demonstrate that HIPL outperforms previous methods in few-shot classification and base-to-new generalization. Our code is available athttps://github.com/Yxt1212/HIPL

Existing monocular depth estimation methods have achieved satisfactory performance on wild datasets. However, these methods are usually trained and tested on a single dataset, which makes them difficult to generalize to other scenarios. To learn diverse scene priors from multiple datasets, we propose a hierarchical framework with adaptive bins for robust monocular depth estimation, which consists of two critical components: a group-wise query generator to assign hierarchical bins and a correlation-aware transformer decoder to generate adaptive bin features. The proposed HA-Bins enjoys several merits. First, the group-wise query generator progressively increases the number of bin queries for multi-scale image features, resulting in a hierarchical bin distribution robust to diverse scenarios. Second, the correlation-aware transformer decoder refines the correlation of bin queries and image features, effectively improving adaptive image feature aggregation. We visualize the query activation maps on NYUDepthv2 dataset, showing that the proposed network effectively suppresses the depth-irrelevant regions. Experiments on KITTI, Sintel, and RabbitAI benchmarks show that without any fine-tuning, our model jointly trained on multiple datasets achieves competitive performance with the state-of-the-art and solid robustness toward diverse scenarios. In addition, our method wins second place in Robust Vision Challenge 2022 towards challenging scenarios with different characteristics.

Deep neural networks are known to be vulnerable to adversarial examples, where adding carefully crafted adversarial perturbations to the inputs can mislead the DNN model. However, it is challenging to generate effective adversarial examples in the physical world due to many uncontrollable physical dynamics, which pose security and safety threats in the real world. Current physical attack methods aim to generate robust physical adversarial examples by simulating all possible physical dynamics. If attacking a new image or a new DNN model, they require expensive manual efforts for simulating physical dynamics or considerable time for iteratively optimizing. To tackle these limitations, we propose a robust and generalized physical adversarial attack method with Meta-GAN (Meta-GAN Attack), which is able to not only generate robust physical adversarial examples, but also generalize to attacking novel images and novel DNN models by accessing a few digital and physical images. First, we propose to craft robust physical adversarial examples with a generative attack model via simulating color and shape distortions. Second, we formulate the physical attack as a few-shot learning problem and design a novel class-agnostic and model-agnostic meta-learning algorithm to solve this problem. Extensive experiments on two benchmark datasets with four challenging experimental settings verify the superior robustness and generalization of our method by comparing to state-of-the-art physical attack methods. The source code is released at github.

Despite the rapid progress and significant success of deep learning in a wide spectrum of fields, adversarial examples expose many security threats to deep learning models. Recently, an interesting property has been discovered that adversarial examples are transferable, which means adversarial examples targeting a given model can also attack another model. Therefore, many researchers are attracted by this property and work on how to improve the transferability of adversarial examples. Furthermore, compared to the traditional attack methods of disrupting output logits (dubbed logit-based attacks), recent works reveal that disrupting feature maps instead of logits can lead to more transferable adversarial examples (dubbed feature-based attacks). However, previous feature-based attacks mostly hold the intuitive designs of the optimization goals and specialization for certain scenarios with a lack of theoretical motivations and a unified framework. To overcome these limitations, we propose a Unified Feature-based Attack Framework, dubbed as UFAF, combining a dispersion loss and a distance loss, which unifies eight existing feature-based attacks. Furthermore, we also bridge the formulation gap between feature-based attacks and traditional logit-based attacks. With our UFAF, we propose an Entropy-Wasserstein (EW) attack by specifying the dispersion loss as Entropy and the distance loss as Wasserstein Distance, respectively. Besides, we provide theoretical analysis to guarantee the effectiveness of the proposed attack method. Extensive experimental results show the superior performance of our EW attack, which can outperform state-of-the-art attacks by 4.95% on attack success rates in untargeted attack settings, and by 1.95% on targeted transfer rates and 1.17% on target success rates in targeted attack settings. Moreover, our framework can help other feature-based attacks improve their performance by 7.7% in untargeted attack settings.

Recent learning-based methods demonstrate their strong ability to estimate depth for multi-view stereo reconstruction. However, most of these methods directly extract features via regular or deformable convolutions, and few works consider the alignment of the receptive fields between views while constructing the cost volume. Through analyzing the constraint and inference of previous MVS networks, we find that there are still some shortcomings that hinder the performance. To deal with the above issues, we propose an Epipolar-Guided Multi-View Stereo Network with Interval-Aware Label (EI-MVSNet), which includes an epipolar-guided volume construction module and an interval-aware depth estimation module in a unified architecture for MVS. The proposed EI-MVSNet enjoys several merits. First, in the epipolar-guided volume construction module, we construct cost volume with features from aligned receptive fields between different pairs of reference and source images via epipolar-guided convolutions, which take rotation and scale changes into account. Second, in the interval-aware depth estimation module, we attempt to supervise the cost volume directly and make depth estimation independent of extraneous values by perceiving the upper and lower boundaries, which can achieve fine-grained predictions and enhance the reasoning ability of the network. Extensive experimental results on two standard benchmarks demonstrate that our EI-MVSNet performs favorably against state-of-the-art MVS methods. Specifically, our EI-MVSNet ranks$1_{st}$on both intermediate and advanced subsets of the Tanks and Temples benchmark, which verifies the high precision and strong robustness of our model.

We tackle the problem of establishing dense correspondences between a pair of images in an efficient way. Most existing dense matching methods use 4D convolutions to filter incorrect matches, but 4D convolutions are highly inefficient due to their quadratic complexity. Besides, these methods learn features with fixed convolutions which cannot make learnt features robust to different challenge scenarios. To deal with these issues, we propose an Efficient Dynamic Correspondence Network (EDCNet) by jointly equipping pre-separate convolution (Psconv) and dynamic convolution (Dyconv) to establish dense correspondences in a coarse-to-fine manner. The proposed EDCNet enjoys several merits. First, two well-designed modules including a neighbourhood aggregation (NA) module and a dynamic feature learning (DFL) module are combined elegantly in the coarse-to-fine architecture, which is efficient and effective to establish both reliable and accurate correspondences. Second, the proposed NA module maintains linear complexity, showing its high efficiency. And our proposed DFL module has better flexibility to learn features robust to different challenges. Extensive experimental results show that our algorithm performs favorably against state-of-the-art methods on three challenging datasets including HPatches, Aachen Day-Night and InLoc.

The mainstream of image and sentence matching studies currently focuses on fine-grained alignment of image regions and sentence words. However, these methods miss a crucial fact: the correspondence between images and sentences does not simply come from alignments between individual regions and words but from alignments between the phrases they form respectively. In this work, we propose a novel Decoupled Cross-modal Phrase-Attention network (DCPA) for image-sentence matching by modeling the relationships between textual phrases and visual phrases. Furthermore, we design a novel decoupled manner for training and inferencing, which is able to release the trade-off for bi-directional retrieval, where image-to-sentence matching is executed in textual semantic space and sentence-to-image matching is executed in visual semantic space. Extensive experimental results on Flickr30K and MS-COCO demonstrate that the proposed method outperforms state-of-the-art methods by a large margin, and can compete with some methods introducing external knowledge.

Generalized Zero-Shot Learning (GZSL) aims at recognizing images from both seen and unseen classes by constructing correspondences between visual images and semantic embedding. However, existing methods suffer from a strong bias problem, where unseen images in the target domain tend to be recognized as seen classes in the source domain. To address this issue, we propose a Prototype-augmented Self-supervised Generative Network by integrating self-supervised learning and prototype learning into a feature generating model for GZSL. The proposed model enjoys several advantages. First, we propose a Self-supervised Learning Module to exploit inter-domain relationships, where we introduce anchors as a bridge between seen and unseen categories. In the shared space, we pull the distribution of the target domain away from the source domain and obtain domain-aware features. To our best knowledge, this is the first work to introduce self-supervised learning into GZSL as learning guidance. Second, a Prototype Enhancing Module is proposed to utilize class prototypes to model reliable target domain distribution in finer granularity. In this module, a Prototype Alignment mechanism and a Prototype Dispersion mechanism are combined to guide the generation of better target class features with intra-class compactness and inter-class separability. Extensive experimental results on five standard benchmarks demonstrate that our model performs favorably against state-of-the-art GZSL methods.

Most existing tracking methods try to represent the target by exploiting visual information as much as possible based on the various deep networks. However, the appearance model hardly describes the attribute feature of the target well, which makes the trackers fail to adapt to the complex visual surrounding. In this article, inspired by brain-like intelligence, we propose an One-stream Vision-Language Memory network (OVLM) for object tracking. Firstly, we use the combination of vision and language to build the target model and use the semantic information in the language to compensate for the instability of visual information, making the target model more stable in the face of complex appearance changes. Secondly, to build a more compact target model, we propose a memory token selection mechanism that utilizes linguistic information to eliminate tokens that do not contain target information. Furthermore, to provide better visual information for target modeling, we propose a language-based evaluation method to select high-quality target samples to be stored in the memory. Finally, OVLM achieves a 64.7% success rate on the large-scale tracking benchmark dataset TNL2K, outperforming the previous best result (VLT) by 11.6%. By exposing the possibility of the vision-language memory network, we aim to draw greater attention to it and open up new avenues for vision-language tracking.

Recently, template-based trackers have become the leading tracking algorithms with promising performance in terms of efficiency and accuracy. However, the correlation operation between query feature and the given template only achieves accurate target localization, but is prone to state estimation error, especially when the target suffers from severe deformation. To address this issue, segmentation-based trackers are proposed that use per-pixel matching to improve the tracking performance of deformable objects effectively. However, most of the existing trackers only match with the target features of the initial frame, thereby lacking the discrimination for handling a variety of challenging factors, e.g., similar distractors, background clutter, and appearance change. To this end, we propose a dynamic compact memory embedding technique to enhance the discrimination of the segmentation-based visual tracking method that can well tell the target from the background. Specifically, we initialize a memory embedding with the target features in the first frame. During the tracking process, the current target features that have certain correlation with the existing memory are updated to the memory embedding online. To further improve the tracking accuracy for deformable objects, we use a weighted point-to-global matching strategy to measure the correlation between the pixelwise query feature and the whole template, so as to capture more detailed deformation information. Extensive evaluations on six challenging tracking benchmarks including VOT2016, VOT2018, VOT2019, GOT-10K, TrackingNet, and LaSOT demonstrate the superiority of our method over recent remarkable trackers. Besides, our tracker outperforms the excellent segmentation-based trackers, i.e., D3S and SiamMask on the DAVIS2017 benchmark. The code is available at https://github.com/peace-love243/CMEDFL.

A memory mechanism has attracted growing popularity in tracking tasks due to the ability of learning long-term-dependent information. However, it is very challenging for existing memory modules to provide the intrinsic attribute information of the target to the tracker in complex scenes. In this article, by considering the biological visual memory mechanisms, we propose the novel online tracking method via an attention-driven memory network, which can mine discriminative memory information and enhance the robustness and reliability of the tracker. First, to reinforce effectiveness of memory content, we design a novel attention-driven memory network. In the network, the long memory module gains property-level memory information by focusing on the state of the target at both the channel and spatial levels. Meanwhile, in reciprocity, we add a short-term memory module to maintain good adaptability when confronting drastic deformation of the target. The attention-driven memory network can adaptively adjust the contribution of short-term and long-term memories to tracking results under the weighted gradient harmonized loss. On this basis, to avoid model performance degradation, an online memory updater (MU) is further proposed. It is designed to mining for target information in tracking results through the Mixer layer and the online head network together. By evaluating the confidence of the tracking results, the memory updater can accurately judge the time of updating the model, which guarantees the effectiveness of online memory updates. Finally, the proposed method performs favorably and has been extensively validated on several benchmark datasets, including object tracking benchmark-50/100 (OTB-50/100), temple color-128 (TC-128), unmanned aerial vehicles-123 (UAV-123), generic object tracking -10k (GOT-10k), visual object tracking-2016 (VOT-2016), and VOT-2018 against several advanced methods.

To reduce computational redundancies, a common approach is to integrate detection and re-identification (Re-ID) into a single network in multi-object tracking (MOT), referred to as “tracking by detection.” Most of the previous research has focused on resolving the conflict between the detection and Re-ID branches, considering it a simple coupling. In our work, we uncover that the entangled state between the detection and Re-ID tasks is much more complex than previous idea, resulting in a form of competition that degrades performance. To address the preceding issue, we propose a feature disentanglement network that deeply disentangles the intricately interwoven latent space of features and provides differentiated feature maps for each individual task. Furthermore, considering the demand for shallow semantic features in the feature re-ID branch, we also introduce a feature re-globalization module to enrich the shallow semantics. By integrating two distinct networks into a one-shot online MOT method, we develop a robust MOT tracker (named HDGTrack ). We conduct extensive experiments on a number of benchmarks, and our experimental results demonstrate that our method significantly outperforms state-of-the-art MOT methods. Besides, HDGTrack is efficient and can run at 13.9 (MOT17) and 8.7 (MOT20) frames per second.

Recently, deep Stereo Matching (SM) networks have shown impressive performance and attracted increasing attention in computer vision. However, existing deep SM networks are prone to learn dataset-dependent shortcuts, which fail to generalize well on unseen realistic datasets. This paper takes a step towards training robust models for the domain generalized SM task, which mainly focuses on learning shortcut-invariant representation from synthetic data to alleviate the domain shifts. Specifically, we propose a Hierarchical Visual Transformation (HVT) network to 1) first transform the training sample hierarchically into new domains with diverse distributions from three levels: Global, Local, and Pixel, 2) then maximize the visual discrepancy between the source domain and new domains, and minimize the cross-domain feature inconsistency to capture domain-invariant features. In this way, we can prevent the model from exploiting the artifacts of synthetic stereo images as shortcut features, thereby estimating the disparity maps more effectively based on the learned robust and shortcut-invariant representation. We integrate our proposed HVT network with SOTA SM networks and evaluate its effectiveness on several public SM benchmark datasets. Extensive experiments clearly show that the HVT network can substantially enhance the performance of existing SM networks in synthetic-to-realistic domain generalization.

Unsupervised point cloud shape correspondence aims to obtain dense point-to-point correspondences between point clouds without manually annotated pairs. However, humans and some animals have bilateral symmetry and various orientations, which lead to severe mispredictions of symmetrical parts. Besides, point cloud noise disrupts consistent representations for point cloud and thus degrades the shape correspondence accuracy. To address the above issues, we propose a Self-Ensembling ORientation-aware Network termed SE-ORNet. The key of our approach is to exploit an orientation estimation module with a domain adaptive discriminator to align the orientations of point cloud pairs, which significantly alleviates the mispredictions of symmetrical parts. Additionally, we design a self-ensembling framework for unsupervised point cloud shape correspondence. In this framework, the disturbances of point cloud noise are overcome by perturbing the inputs of the student and teacher networks with different data augmentations and constraining the consistency of predictions. Extensive experiments on both human and animal datasets show that our SE-ORNet can surpass state-of-the-art unsupervised point cloud shape correspondence methods.

Adversarial attacks can evaluate model robustness and have been of great concern in recent years. Among various attacks, targeted attacks aim at misleading victim models to output adversary-desired predictions, which are more challenging and threatening than untargeted ones. Existing targeted attacks can be roughly divided into instance-specific and instance-agnostic attacks. Instance-specific attacks craft adversarial examples via iterative gradient updating on the specific instance. In contrast, instance-agnostic attacks learn a universal perturbation or a generative model on the global dataset to perform attacks. However, they rely too much on the classification boundary of substitute models, ignoring the realistic distribution of the target class, which may result in limited targeted attack performance. And there is no attempt to simultaneously combine the information of the specific instance and the global dataset. To deal with these limitations, we first conduct an analysis via a causal graph and propose to craft transferable targeted adversarial examples by injecting target patterns. Based on this analysis, we introduce a generative attack model composed of a cross-attention guided convolution module and a pattern injection module. Concretely, the former adopts a dynamic convolution kernel and a static convolution kernel for the specific instance and the global dataset, respectively, which can inherit the advantages of both instance-specific and instance-agnostic attacks. And the pattern injection module utilizes a pattern prototype to encode target patterns, which can guide the generation of targeted adversa rial examples. Besides, we also provide rigorous theoretical analysis to guarantee the effectiveness of our method. Extensive experiments demonstrate that our method shows superior performance than 10 existing adversarial attacks against 13 models.

Establishing pixel-level matches between image pairs is vital for a variety of computer vision applications. How-ever, achieving robust image matching remains challenging because CNN extracted descriptors usually lack discrim-inative ability in texture-less regions and keypoint detec-tors are only good at identifying keypoints with a specific level of structure. To deal with these issues, a novel im-age matching method is proposed by Jointly Learning Hier-archical Detectors and Contextual Descriptors via Agent-based Transformers (D2Former), including a contextual feature descriptor learning (CFDL) module and a hierar-chical keypoint detector learning (HKDL) module. The proposed D2Former enjoys several merits. First, the pro-posed CFDL module can model long-range contexts effi-ciently and effectively with the aid of designed descriptor agents. Second, the HKDL module can generate keypoint detectors in a hierarchical way, which is helpful for detecting keypoints with diverse levels of structures. Extensive experimental results on four challenging benchmarks show that our proposed method significantly outperforms state-of-the-art image matching methods.

Camouflaged Instance Segmentation (CIS) aims at predicting the instance-level masks of camouflaged objects, which are usually the animals in the wild adapting their appearance to match the surroundings. Previous instance segmentation methods perform poorly on this task as they are easily disturbed by the deceptive camouflage. To address these challenges, we propose a novel De-camouflaging Network (DCNet) including a pixel-level camouflage decoupling module and an instance-level camouflage suppression module. The proposed DCNet enjoys several merits. First, the pixel-level camouflage decoupling module can extract camouflage characteristics based on the Fourier transformation. Then a difference attention mechanism is proposed to eliminate the camouflage characteristics while reserving target object characteristics in the pixel feature. Second, the instance-level camouflage suppression module can aggregate rich instance information from pixels by use of instance prototypes. To mitigate the effect of background noise during segmentation, we introduce some reliable reference points to build a more robust similarity measurement. With the aid of these two modules, our DCNet can effectively model de-camouflaging and achieve accurate segmentation for camouflaged instances. Extensive experimental results on two benchmarks demonstrate that our DCNet performs favorably against state-of-the-art CIS methods, e.g., with more than 5% performance gains on COD10K and NC4K datasets in average precision.

Automatic mitochondria segmentation enjoys great popularity with the development of deep learning. However, existing methods rely heavily on the labor-intensive manual gathering by experienced domain experts. And naively applying semi-supervised segmentation methods in the natural image field to mitigate the labeling cost is undesirable. In this work, we analyze the gap between mitochondrial images and natural images and rethink how to achieve effective semi-supervised mitochondria segmentation, from the perspective of reliable prototype-level supervision. We propose a novel end-to-end dual-reliable (DualRel) network, including a reliable pixel aggregation module and a reliable prototype selection module. The proposed DualRel enjoys several merits. First, to learn the prototypes well without any explicit supervision, we carefully design the referential correlation to rectify the direct pairwise correlation. Second, the reliable prototype selection module is responsible for further evaluating the reliability of prototypes in constructing prototype-level consistency regularization. Extensive experimental results on three challenging benchmarks demonstrate that our method performs favorably against state-of-the-art semi-supervised segmentation methods. Importantly, with extremely few samples used for training, DualRel is also on par with current state-of-the-art fully supervised methods.

Weakly-supervised temporal action localization aims to localize and recognize actions in untrimmed videos with only video-level category labels during training. Without instance-level annotations, most existing methods follow the Segment-based Multiple Instance Learning (S-MIL) framework, where the predictions of segments are supervised by the labels of videos. However, the objective for acquiring segment-level scores during training is not consistent with the target for acquiring proposal-level scores during testing, leading to suboptimal results. To deal with this problem, we propose a novel Proposal-based Multiple Instance Learning (P-MIL) framework that directly classifies the candidate proposals in both the training and testing stages, which includes three key designs: 1) a surrounding contrastive feature extraction module to suppress the discriminative short proposals by considering the surrounding contrastive information, 2) a proposal completeness evaluation module to inhibit the low-quality proposals with the guidance of the completeness pseudo labels, and 3) an instance-level rank consistency loss to achieve robust detection by leveraging the complementarity of RGB and FLOW modalities. Extensive experimental results on two challenging benchmarks including THUMOS14 and ActivityNet demonstrate the superior performance of our method. Our code is available at github.com/RenHuan1999/CVPR2023_P-MIL.

Few-shot segmentation (FSS) aims to segment novel ob-jects in a given query image with only a few annotated support images. However, most previous best-performing methods, whether prototypical learning methods or affinity learning methods, neglect to alleviate false matches caused by their own pixel-level correlation. In this work, we rethink how to mitigate the false matches from the perspective of representative reference features (referred to as buoys), and propose a novel adaptive buoys correlation (ABC) network to rectify direct pairwise pixel-level correlation, including a buoys mining module and an adaptive correlation module. The proposed ABC enjoys several merits. First, to learn the buoys well without any correspondence supervision, we customize the buoys mining module according to the three characteristics of representativeness, task awareness and re-silience. Second, the proposed adaptive correlation module is responsible for further endowing buoy-correlation-based pixel matching with an adaptive ability. Extensive experimen-tal results with two different backbones on two challenging benchmarks demonstrate that our ABC, as a general plu-gin, achieves consistent improvements over several leading methods on both I-shot and 5-shot settings.

Local feature matching aims at finding correspondences between a pair of images. Although current detector-free methods leverage Transformer architecture to obtain an impressive performance, few works consider maintaining local consistency. Meanwhile, most methods struggle with large scale variations. To deal with the above issues, we propose Adaptive Spot-Guided Transformer (ASTR) for local feature matching, which jointly models the local consistency and scale variations in a unified coarse-to-fine architecture. The proposed ASTR enjoys several merits. First, we design a spot-guided aggregation module to avoid interfering with irrelevant areas during feature aggregation. Second, we design an adaptive scaling module to adjust the size of grids according to the calculated depth information at fine stage. Extensive experimental results on five standard benchmarks demonstrate that our ASTR performs favorably against state-of-the-art methods. Our code will be released on https://astr2023.github.io.

3D instance segmentation aims to predict a set of object instances in a scene and represent them as binary foreground masks with corresponding semantic labels. However, object instances are diverse in shape and category, and point clouds are usually sparse, unordered, and irregular, which leads to a query sampling dilemma. Besides, noise background queries interfere with proper scene perception and accurate instance segmentation. To address the above issues, we propose the Query Refinement Transformer termed QueryFormer. The key to our approach is to exploit a query initialization module to optimize the initialization process for the query distribution with a high coverage and low repetition rate. Additionally, we design an affiliated transformer decoder that suppresses the interference of noise background queries and helps the foreground queries focus on instance discriminative parts to predict final segmentation results. Extensive experiments on ScanNetV2 and S3DIS datasets show that our QueryFormer can surpass state-of-the-art 3D instance segmentation methods.

Mitochondria, as tiny structures within the cell, are of significant importance in studying cell functions for biological and clinical analysis. And exploring how to automatically segment mitochondria in electron microscopy (EM) images has attracted increasing attention. However, most of existing methods struggle to adapt to different scales and appearances of the input due to the inherent limitations of the traditional CNN architecture. To mitigate these limitations, we propose a novel adaptive template transformer (ATFormer) for mitochondria segmentation. The proposed ATFormer model enjoys several merits. First, the designed structural template learning module can acquire appearance-adaptive templates of background, foreground and contour to sense the characteristics of different shapes of mitochondria. And we further adopt an optimal transport algorithm to enlarge the discrepancy among diverse templates to activate corresponding regions fully. Second, we introduce a hierarchical attention learning mechanism to absorb multi-level information for templates to be adaptive scale-aware classifiers for dense prediction. Extensive experimental results on three challenging benchmarks including MitoEM, Lucchi and NucMM-Z datasets demonstrate that our ATFormer performs favorably against state-of-the-art mitochondria segmentation methods.

Memory-based methods in semi-supervised video object segmentation task achieve competitive performance by performing dense matching between query and memory frames. However, most of the existing methods neglect the fact that videos carry rich temporal information yet redundant spatial information. In this case, direct pixel-level global matching will lead to ambiguous correspondences. In this work, we reconcile the inherent tension of spatial and temporal information to retrieve memory frame information along the object trajectory, and propose a novel and coherent Trajectory Memory Retrieval Network (TMRN) to equip with the trajectory information, including a spatial alignment module and a temporal aggregation module. The proposed TMRN enjoys several merits. First, TMRN is empowered to characterize the temporal correspondence which is in line with the nature of video in a data-driven manner. Second, we elegantly customize the spatial alignment module by coupling SVD initialization with agent-level correlation for representative agent construction and rectifying false matches caused by direct pairwise pixel-level correlation, respectively. Extensive experimental results on challenging benchmarks including DAVIS 2017 validation / test and Youtube-VOS 2018/2019 demonstrate that our TMRN, as a general plugin module, achieves consistent improvements over several leading methods.

Semi-supervised 3D object detection from point cloud aims to train a detector with a small number of labeled data and a large number of unlabeled data. The core of existing methods lies in how to select high-quality pseudo-labels using the designed quality evaluation criterion. However, these methods treat each pseudo bounding box as a whole and assign equal importance to each side during training, which is detrimental to model performance due to many sides having poor localization quality. Besides, existing methods filter out a large number of low-quality pseudo-labels, which also contain some correct regression values that can help with model training. To address the above issues, we propose a side-aware framework for semi-supervised 3D object detection consisting of three key designs: a 3D bounding box parameterization method, an uncertainty estimation module, and a pseudo-label selection strategy. These modules work together to explicitly estimate the localization quality of each side and assign different levels of importance during the training phase. Extensive experiment results demonstrate that the proposed method can consistently outperform baseline models under different scenes and evaluation criteria. Moreover, our method achieves state-of-the-art performance on three datasets with different labeled ratios.

Scene boundary detection breaks down long videos into meaningful story-telling units and plays a crucial role in high-level video understanding. Despite significant advancements in this area, this task remains a challenging problem as it requires a comprehensive understanding of multimodal cues and high-level semantics. To tackle this issue, we propose a multimodal high-order relation transformer, which integrates a high-order encoder and an adaptive decoder in a unified framework. By modeling the mul-timodal cues and exploring similarities between the shots, the encoder is capable of capturing high-order relations between shots and extracting shot features with context semantics. By clustering the shots adaptively, the decoder can discover more universal switch pattern between successive scenes, thus helping scene boundary detection. Extensive experimental results on three standard benchmarks demonstrate that the proposed model performs favorably against state-of-the-art video scene detection methods.

Visual object tracking is a fundamental research topic with a broad range of applications. Benefiting from the rapid development of Transformer, pure Transformer trackers have achieved great progress. However, the feature learning of these Transformer-based trackers is easily disturbed by complex backgrounds. To address the above limitations, we propose a novel foreground-background distribution modeling transformer for visual object tracking (F-BDMTrack), including a fore-background agent learning (FBAL) module and a distribution-aware attention (DA2) module in a unified transformer architecture. The proposed F-BDMTrack enjoys several merits. First, the proposed FBAL module can effectively mine fore-background information with designed fore-background agents. Second, the DA2module can suppress the incorrect interaction between foreground and background by modeling fore-background distribution similarities. Finally, F-BDMTrack can extract discriminative features under ever-changing tracking scenarios for more accurate target state estimation. Extensive experiments show that our F-BDMTrack outperforms previous state-of-the-art trackers on eight tracking benchmarks.

Neural connectivity reconstruction aims to understand the function of biological reconstruction and promote basic scientific research. The intricate morphology and densely intertwined branches make it an extremely challenging task. Most previous best-performing methods adopt affinity learning or metric learning. Nevertheless, they either neglect to model explicit voxel semantics caused by implicit optimization or are hysteresis to spatial information. Furthermore, the inherent locality of 3D CNNs limits modeling long-range dependencies, leading to sub-optimal results. In this work, we propose a coherent and unified Appearance Prompt Vision Transformer (APViT) to integrate affinity and metric learning to exploit the complementarity by learning long-range spatial dependencies. The proposed APViT enjoys several merits. First, the extension continuity-aware attention module aims at constructing hierarchical attention customized for neuron extensibility and slice continuity to learn instance voxel semantic context from a global perspective and utilize continuity priors to enhance voxel spatial awareness. Second, the appearance prompt modulator is responsible for leveraging voxel-adaptive appearance knowledge conditioned on affinity rich in spatial information to instruct instance voxel semantics, exploiting the potential of affinity learning to complement metric learning. Extensive experimental results on multiple challenging benchmarks demonstrate that our APViT achieves consistent improvements with huge flexibility under the same post-processing strategy.

The critical challenge of semi-supervised semantic segmentation lies in how to fully exploit a large volume of unlabeled data to improve the model’s generalization performance for robust segmentation. Existing methods tend to employ certain criteria (weighting function) to select pixel-level pseudo labels. However, the trade-off exists between inaccurate yet utilized pseudo-labels, and correct yet discarded pseudo-labels in these methods when handling pseudo-labels without thoughtful consideration of the weighting function, hindering the generalization ability of the model. In this paper, we systematically analyze the trade-off in previous methods when dealing with pseudo-labels. We formally define the trade-off between inaccurate yet utilized pseudo-labels, and correct yet discarded pseudo-labels by explicitly modeling the confidence distribution of correct and inaccurate pseudo-labels, equipped with a unified weighting function. To this end, we propose Distribution-Aware Weighting (DAW) to strive to minimize the negative equivalence impact raised by the trade-off. We find an interesting fact that the optimal solution for the weighting function is a hard step function, with the jump point located at the intersection of the two confidence distributions. Besides, we devise distribution alignment to mitigate the issue of the discrepancy between the prediction distributions of labeled and unlabeled data. Extensive experimental results on multiple benchmarks including mitochondria segmentation demonstrate that DAW performs favorably against state-of-the-art methods.

Few-shot segmentation (FSS) aims to segment objects of new categories given only a handful of annotated samples. Previous works focus their efforts on exploring the support information while paying less attention to the mining of the critical query branch. In this paper, we rethink the importance of support information and propose a new query-centric FSS model Adversarial Mining Transformer (AMFormer), which achieves accurate query image segmentation with only rough support guidance or even weak support labels. The proposed AMFormer enjoys several merits. First, we design an object mining transformer (G) that can achieve the expansion of incomplete region activated by support clue, and a detail mining transformer (D) to discriminate the detailed local difference between the expanded mask and the ground truth. Second, we propose to train G and D via an adversarial process, where G is optimized to generate more accurate masks approaching ground truth to fool D. We conduct extensive experiments on commonly used Pascal-5i and COCO-20i benchmarks and achieve state-of-the-art results across all settings. In addition, the decent performance with weak support labels in our query-centric paradigm may inspire the development of more general FSS models.

Establishing effective correspondences between a pair of images is difficult due to real-world challenges such as illumination, viewpoint and scale variations. Modern detector-based methods typically learn fixed detectors from a given dataset, which is hard to extract repeatable and reliable keypoints for various images with extreme appearance changes and weakly textured scenes. To deal with this problem, we propose a novel Dynamic Keypoint Detection Network (DKDNet) for robust image matching via a dynamic keypoint feature learning module and a guided heatmap activator. The proposed DKDNet enjoys several merits. First, the proposed dynamic keypoint feature learning module can generate adaptive keypoint features via the attention mechanism, which is flexibly updated with the current input image and can capture keypoints with different patterns. Second, the guided heatmap activator can effectively fuse multi-group keypoint heatmaps by fully considering the importance of different feature channels, which can realize more robust keypoint detection. Extensive experimental results on four standard benchmarks demonstrate that our DKDNet outperforms state-of-the-art image-matching methods by a large margin. Specifically, our DKDNet can outperform the best image-matching method by 2.1% in AUC@ 3px on HPatches, 3.74% in AUC@$5^\circ$on ScanNet, 7.14% in AUC@$5^\circ$on MegaDepth and 12.32% in AUC@$5^\circ$on YFCC100M.

Visual tracking aims to estimate object state in a video sequence, which is challenging when facing drastic appearance changes. Most existing trackers conduct tracking with divided parts to handle appearance variations. However, these trackers commonly divide target objects into regular patches by a hand-designed splitting way, which is too coarse to align object parts well. Besides, a fixed part detector is difficult to partition targets with arbitrary categories and deformations. To address the above issues, we propose a novel adaptive part mining tracker (APMT) for robust tracking via a transformer architecture, including an object representation encoder, an adaptive part mining decoder, and an object state estimation decoder. The proposed APMT enjoys several merits. First, in the object representation encoder, object representation is learned by distinguishing target object from background regions. Second, in the adaptive part mining decoder, we introduce multiple part prototypes to adaptively capture target parts through cross-attention mechanisms for arbitrary categories and deformations. Third, in the object state estimation decoder, we propose two novel strategies to effectively handle appearance variations and distractors. Extensive experimental results demonstrate that our APMT achieves promising results with high FPS. Notably, our tracker is ranked the first place in the VOT-STb2022 challenge.

Weakly supervised object localization (WSOL) aims to predict both object locations and categories with only image-level class labels. However, most existing methods rely on class-specific image regions for localization, resulting in incomplete object localization. To alleviate this problem, we propose a novel end-to-end task-aware framework with a transformer encoder-decoder architecture (TAFormer) to learn class-agnostic foreground maps, including a representation encoder, a localization decoder, and a classification decoder. The proposed TAFormer enjoys several merits. First, the designed three modules can effectively perform class-agnostic localization and classification in a task-aware manner, achieving remarkable performance for both tasks. Second, an optimal transport algorithm is proposed to provide pixel-level pseudo labels to online refine foreground maps. To the best of our knowledge, this is the first work by exploring a task-aware framework with a transformer architecture and an optimal transport algorithm to achieve accurate object localization for WSOL. Extensive experiments with four backbones on two standard benchmarks demonstrate that our TAFormer achieves favorable performance against state-of-the-art methods. Furthermore, we show that the proposed TAFormer provides higher robustness against adversarial attacks and noisy labels.

In weakly supervised (WSAL) and unsupervised temporal action localization (UAL), the target is to simultaneously localize temporal boundaries and identify category labels of actions with only video-level category labels (WSAL) or category numbers in a dataset (UAL) during training. Among existing methods, attention based methods have achieved superior performance in both tasks by highlighting action segments with foreground attention weights. However, without the segment-level supervision on the attention weight learning, the quality of the attention weight hinders the performance of these methods. In this paper, we propose a novel Uncertainty Guided Collaborative Training (UGCT) strategy to alleviate this problem, which mainly includes two key designs: (1) The first design is an online pseudo label generation module, in which the RGB and FLOW streams work collaboratively to learn from each other. (2) The second design is an uncertainty aware learning module, which can mitigate the noise in the generated pseudo labels. These two designs work together to promote the model performance effectively and efficiently by exchanging information between RGB and FLOW streams. Extensive experimental results on two benchmark datasets with three attention based methods demonstrate the effectiveness of the proposed method, e.g, more than 7.0% performance gain for mAP@IoU=0.5 on THUMOS14 dataset.

3D object detection in point cloud aims at simultaneously localizing and recognizing 3D objects from a 3D point set. However, since point clouds are usually sparse, unordered, and irregular, it is challenging to learn robust point representations and sample high-quality object queries. To deal with the above issues, we propose a Long-short rangE Adaptive transformer with Dynamic sampling (LeadNet), including a point representation encoder, a dynamic object query sampling decoder, and an object detection decoder in a unified architecture for 3D object detection. Specifically, in the point representation encoder, we combine an attention layer and a channel attentive kernel convolution layer to consider the local structure and the long-range context simultaneously. In the dynamic object query sampling decoder, we utilize multiple dynamic prototypes to adapt to various point clouds. In the object detection decoder, we incorporate a dynamic Gaussian weight map into the cross-attention mechanism to help the detection decoder focus on the proper visual regions near the object, further accelerating the training process. Extensive experimental results on two standard benchmarks show that our LeadNet outperforms the 3DETR baseline by$11.6\%~\text {mAP}_{50}$on the ScanNet v2 dataset and achieves the new state-of-the-art results on ScanNet v2 and SUN RGB-D benchmarks for the geometric-only approaches.

Point cloud shape correspondence aims at accurately mapping one point cloud to another point cloud with various 3D shapes. Since point clouds are usually sparse, disordered, irregular, and with diverse shapes, it is challenging to learn consistent point cloud representations and achieve the accurate matching of different point cloud shapes. To address the above issues, we propose a Hierarchical Shape-consistent TRansformer for unsupervised point cloud shape correspondence (HSTR), including a multi-receptive-field point representation encoder and a shape-consistent constrained module in a unified architecture. The proposed HSTR enjoys several merits. In the multi-receptive-field point representation encoder, we set progressively larger receptive fields in different blocks to simultaneously consider the local structure and the long-range context. In the shape-consistent constrained module, we design two novel shape selective whitening losses, which can complement each other to achieve suppression of features sensitive to shape change. Extensive experimental results on four standard benchmarks demonstrate the superiority and generalization ability of our approach to existing methods at the similar model scale, and our method achieves the new state-of-the-art results.

Lane detection is a challenging task that requires predicting complex topology shapes of lane lines and distinguishing different types of lanes simultaneously. Earlier works follow a top-down roadmap to regress predefined anchors into various shapes of lane lines, which lacks enough flexibility to fit complex shapes of lanes due to the fixed anchor shapes. Lately, some works propose to formulate lane detection as a keypoint estimation problem to describe the shapes of lane lines more flexibly and gradually group adjacent keypoints belonging to the same lane line in a point-by-point manner, which is inefficient and time-consuming during postprocessing. In this paper, we propose a Global Association Network (GANet) to formulate the lane detection problem from a new perspective, where each keypoint is directly regressed to the starting point of the lane line instead of point-by-point extension. Concretely, the association of keypoints to their belonged lane line is conducted by predicting their offsets to the corresponding starting points of lanes globally without dependence on each other, which could be done in parallel to greatly improve efficiency. In addition, we further propose a Lane-aware Feature Aggregator (LFA), which adaptively captures the local correlations between adjacent keypoints to supplement local information to the global association. Extensive experiments on two popular lane detection benchmarks show that our method outperforms previous methods with F1 score of 79.63% on CULane and 97.71% on Tusimple dataset with high FPS.

While the majority of FSL models focus on image classification, the extension to action recognition is rather challenging due to the additional temporal dimension in videos. To address this issue, we propose an end-to-end Motion-modulated Temporal Fragment Alignment Network (MT-FAN) by jointly exploring the task-specific motion modulation and the multi-level temporal fragment alignment for Few-Shot Action Recognition (FSAR). The proposed MT-FAN model enjoys several merits. First, we design a motion modulator conditioned on the learned task-specific motion embeddings, which can activate the channels related to the task-shared motion patterns for each frame. Second, a segment attention mechanism is proposed to automatically discover the higher-level segments for multi-level temporal fragment alignment, which encompasses the frame-to-frame, segment-to-segment, and segment-to-frame alignments. To the best of our knowledge, this is the first work to exploit task-specific motion modulation for FSAR. Extensive experimental results on four standard benchmarks demonstrate that the proposed model performs favorably against the state-of-the-art FSAR methods.

Correlation filters (CF), a popular tool for visual tracking, suffer from unwanted boundary effects due to the periodic assumption needed for FFT implementation. To address this issue, spatially regularized discriminative correlation filters (SRDCF) have been proposed by introducing a weighting matrix to the regularization term. However, the existing design of spatial weighting matrix is often heuristic and non-adaptive. Inspired by recent advances in joint discrimination and reliability learning for correlation tracking, we propose a principled Bayesian correlation filter learning method using Gaussian scale mixture (GSM) model. The key idea is to decompose each CF coefficient into the product of a positive scalar multiplier and a Gaussian random variable. Treating positive multipliers as weighting coefficients, GSM-based modeling of CFs leads to a spatially adaptive regularization strategy with improved capability of handling various appearance-related uncertainty factors (e.g., scale variation, out-of-plane rotation, and motion blur). Moreover, by imposing a sparse prior over the multipliers, we can jointly learn multipliers and CFs under a unified Bayesian estimation framework. Structured GSM model allows us to better exploit the spatial correlations among CFs and further improve the tracking performance. Experimental results on OTB-2013, OTB-2015, Temple Color-128, VOT-2016, and VOT-2017 show that our tracking method performs favorably when compared with current state-of-the-art methods.

Numerous tracking approaches attempt to improve target representation through target-aware or distractor-aware. However, the unbalanced considerations of target or distractor information make it diffcult for these methods to benefit from the two aspects at the same time. In this paper, we propose a target-distractor aware model with discriminative enhancement learning loss to learn target representation, which can better distinguish the target in complex scenes. Firstly, to enlarge the gap between the target and distractor, we design a discriminative enhancement learning loss. By highlighting the hard negatives that are similar to the target and shrinking the easy negatives that are pure background, the features sensitive to the target or distractor representation can be more conveniently mined. On this basis, we further propose a target-distractor aware model. Unlike existing methods of preference target or distractor, we construct the target-specific feature space by activating the target-sensitive and the distractor-silence feature. Therefore, the appearance model can not only represent the target well but also suppress the background distractor. Finally, the target-distractor aware target representation model is integrated with a Siamese matching network for visual tracking for achieving robust and realtime visual tracking. Extensive experiments are performed on eight tracking benchmarks show that the proposed algorithm achieves favorable performance.

Temporal and spatial contexts, characterizing target appearance variations and target-background differences, respectively, are crucial for improving the online adaptive ability and instance-level discriminative ability of object tracking. However, most existing trackers focus on either the temporal context or the spatial context during tracking and have not exploited these contexts simultaneously and effectively. In this paper, we propose a Spatial-TEmporal Memory (STEM) network to exploit these contexts jointly for object tracking. Specifically, we develop a key-value structured memory model equipped with a key-value index-based memory reading mechanism to model the spatial and temporal contexts simultaneously. To update the memory with new target states and ensure the diversity of the memory, we introduce a similarity-aware memory update scheme. In addition, we construct an entropy-guided ensemble strategy to fuse the prediction models based on these two contexts, such that these two contexts can be exploited to estimate the target state jointly. Extensive experimental results on eight challenging datasets, including OTB2015, TC128, UAV123, VOT2018, LaSOT, TrackingNet, GOT-10k, and OxUvA, demonstrate that the proposed method performs favorably against state-of-the-art trackers.

Diabetic retinopathy (DR) grading from fundus images has attracted increasing interest in both academic and industrial communities. Most convolutional neural network-based algorithms treat DR grading as a classification task via image-level annotations. However, these algorithms have not fully explored the valuable information in the DR-related lesions. In this article, we present a robust framework, which collaboratively utilizes patch-level and image-level annotations, for DR severity grading. By an end-to-end optimization, this framework can bidirectionally exchange the fine-grained lesion and image-level grade information. As a result, it exploits more discriminative features for DR grading. The proposed framework shows better performance than the recent state-of-the-art algorithms and three clinical ophthalmologists with over nine years of experience. By testing on datasets of different distributions (such as label and camera), we prove that our algorithm is robust when facing image quality and distribution variations that commonly exist in real-world practice. We inspect the proposed framework through extensive ablation studies to indicate the effectiveness and necessity of each motivation. The code and some valuable annotations are now publicly available.

Visible-infrared person re-identification (VI-ReID) is challenging due to the large modality discrepancy between visible and infrared images. Existing methods mainly focus on learning modality-shared representations by embedding images from different modalities into a common feature space, in which some discriminative modality information is discarded. Different from these methods, in this paper, we propose a novel Modality-Specific Memory Network (MSMNet) to complete the missing modality information and aggregate visible and infrared modality features into a unified feature space for the VI-ReID task. The proposed model enjoys several merits. First, it can exploit the missing modality information to alleviate the modality discrepancy when only the single-modality input is provided. To the best of our knowledge, this is the first work to exploit the missing modality information completion and alleviate the modality discrepancy with the memory network. Second, to guide the learning process of the memory network, we design three effective learning strategies, including feature consistency, memory representativeness and structural alignment. By incorporating these learning strategies in a unified model, the memory network can be well learned to propagate identity-related information between modalities and boost the VI-ReID performance. Extensive experimental results on two standard benchmarks (SYSU-MM01 and RegDB) demonstrate that the proposed MSMNet performs favorably against state-of-the-art methods.

Weakly Supervised Object Localization (WSOL) aims to localize objects with only image-level labels, which has better scalability and practicability than fully supervised methods in the actual deployment. However, a common limitation for available techniques based on classification networks is that they only highlight the most discriminative part of the object, not the entire object. To alleviate this problem, we propose a novel end-to-end part discovery model (PDM) to learn multiple discriminative object parts in a unified network for accurate object localization and classification. The proposed PDM enjoys several merits. First, to the best of our knowledge, it is the first work to directly model diverse and robust object parts by exploiting part diversity, compactness, and importance jointly for WSOL. Second, three effective mechanisms including diversity, compactness, and importance learning mechanisms are designed to learn robust object parts. Therefore, our model can exploit complementary spatial information and local details from the learned object parts, which help to produce precise bounding boxes and discriminate different object categories. Extensive experiments on two standard benchmarks demonstrate that our PDM performs favorably against state-of-the-art WSOL approaches.

Weakly supervised object localization (WSOL) aims at localizing objects with only image-level labels, which has better scalability and practicability than fully supervised methods. However, without pixel-level supervision, existing methods tend to generate rough localization maps, which hinders localization performance. To alleviate this problem, we propose an adversarial transformer network (ATNet), which aims to obtain a well-learned localization model with pixel-level pseudo labels. The proposed ATNet enjoys several merits. First, we design an object transformer ( G ) that can generate localization maps and pseudo labels effectively and dynamically, and a part transformer ( D ) to accurately discriminate detailed local differences between localization maps and pseudo labels. Second, we propose to train G and D via an adversarial process, where G can generate more accurate localization maps approaching pseudo labels to fool D . To the best of our knowledge, this is the first work to explore transformers with adversarial training to obtain a well-learned localization model for WSOL. Extensive experiments with four backbones on two standard benchmarks demonstrate that our ATNet achieves favorable performance against state-of-the-art WSOL methods. Besides, our adversarial training can provide higher robustness against adversarial attacks.

The key point in multimodal learning is to learn semantic alignment that finds the correspondence between sub-elements of instances from different modality data. Attention mechanism has shown its power in semantic alignment learning as it enables to densely associate sub-elements across different modalities. However, for each sub-element, existing attention aligns it with all the sub-elements from another modality, while most of them have no correspondence with it, i.e. irrelevant sub-elements. The irrelevant sub-elements will distract the semantic alignment if they are also attended. In this paper, we propose a novel focal attention mechanism to learn more accurate semantic alignment. The focal attention sparsely attends to a subset of sub-elements, which are identified as relevant ones according to their posterior probabilities given each sub-element from another modality. Based on the observation that relevant sub-elements mostly describe the same semantic, the posterior probability can precisely distinguish relevant and irrelevant ones by taking interactions within the same modality into consideration, such that relevant sub-elements get higher and closer posterior probabilities, while irrelevant ones get lower probabilities. Such a design learns better semantic alignment by preventing the interference of irrelevant sub-elements, and it facilitates subsequent multimodal tasks that demand semantic alignment. To validate the effectiveness of the focal attention, we conduct extensive experiments on image-text matching and text-to-image generation, and we propose a bidirectional and stacked version of focal attention for them, respectively. Experimental results on benchmarks show that the focal attention can significantly and consistently outperform state-of-the-arts.

Occluded person re-identification (Re-ID) is a challenging task as persons are frequently occluded by various obstacles or other persons, especially in the crowd scenario. To address these issues, we propose a novel end-to-end Part-Aware Transformer (PAT) for occluded person Re-ID through diverse part discovery via a transformer encoder-decoder architecture, including a pixel context based transformer encoder and a part prototype based transformer decoder. The proposed PAT model enjoys several merits. First, to the best of our knowledge, this is the first work to exploit the transformer encoder-decoder architecture for occluded person Re-ID in a unified deep model. Second, to learn part prototypes well with only identity labels, we design two effective mechanisms including part diversity and part discriminability. Consequently, we can achieve diverse part discovery for occluded person Re-ID in a weakly supervised manner. Extensive experimental results on six challenging benchmarks for three tasks (occluded, partial and holistic Re-ID) demonstrate that our proposed PAT performs favor-ably against stat-of-the-art methods.

Weakly supervised temporal action localization aims to detect and localize actions in untrimmed videos with only video-level labels during training. However, without frame-level annotations, it is challenging to achieve localization completeness and relieve background interference. In this paper, we present an Action Unit Memory Network (AUMN) for weakly supervised temporal action localization, which can mitigate the above two challenges by learning an action unit memory bank. In the proposed AUMN, two attention modules are designed to update the memory bank adaptively and learn action units specific classifiers. Furthermore, three effective mechanisms (diversity, homogeneity and sparsity) are designed to guide the updating of the memory network. To the best of our knowledge, this is the first work to explicitly model the action units with a memory network. Extensive experimental results on two standard benchmarks (THUMOS14 and ActivityNet) demonstrate that our AUMN performs favorably against state-of-the-art methods. Specifically, the average mAP of IoU thresholds from 0.1 to 0.5 on the THUMOS14 dataset is significantly improved from 47.0% to 52.1%.

Diabetic retinopathy (DR) is the leading cause of permanent blindness in the working-age population. And automatic DR diagnosis can assist ophthalmologists to design tailored treatments for patients, including DR grading and lesion discovery. However, most of existing methods treat DR grading and lesion discovery as two independent tasks, which require lesion annotations as a learning guidance and limits the actual deployment. To alleviate this problem, we propose a novel lesion-aware transformer (LAT) for DR grading and lesion discovery jointly in a unified deep model via an encoder-decoder structure including a pixel relation based encoder and a lesion filter based decoder. The proposed LAT enjoys several merits. First, to the best of our knowledge, this is the first work to formulate lesion discovery as a weakly supervised lesion localization problem via a transformer decoder. Second, to learn lesion filters well with only image-level labels, we design two effective mechanisms including lesion region importance and lesion region diversity for identifying diverse lesion regions. Extensive experimental results on three challenging benchmarks including Messidor-1, Messidor-2 and EyePACS demonstrate that the proposed LAT performs favorably against state-of-the-art DR grading and lesion discovery methods.

Weakly supervised temporal action detection aims to localize temporal boundaries of actions and identify their categories simultaneously with only video-level category labels during training. Among existing methods, attention based methods have achieved superior performance by separating action and non-action segments. However, without the segment-level ground-truth supervision, the quality of the attention weight hinders the performance of these methods. To alleviate this problem, we propose a novel Uncertainty Guided Collaborative Training (UGCT) strategy, which mainly includes two key designs: (1) The first design is an online pseudo label generation module, in which the RGB and FLOW streams work collaboratively to learn from each other. (2) The second design is an uncertainty aware learning module, which can mitigate the noise in the generated pseudo labels. These two designs work together to promote the model performance effectively and efficiently by imposing pseudo label supervision on attention weight learning. Experimental results on three state-of-the-art attention based methods demonstrate that the proposed training strategy can significantly improve the performance of these methods, e.g., more than 4% for all three methods in terms of mAP@IoU=0.5 on the THUMOS14 dataset.

Modern deep neural networks are often vulnerable to adversarial examples. Most exist attack methods focus on crafting adversarial examples in the digital domain, while only limited works study physical adversarial attack. However, it is more challenging to generate effective adversarial examples in the physical world due to many uncontrollable physical dynamics. Most current physical attack methods aim to generate robust physical adversarial examples by simulating all possible physical dynamics. When attacking new images or new DNN models, they require expensive manually efforts for simulating physical dynamics and considerable time for iteratively optimizing for each image. To tackle these issues, we propose a class-agnostic and model-agnostic physical adversarial attack model (Meta-Attack), which is able to not only generate robust physical adversarial examples by simulating color and shape distortions, but also generalize to attacking novel images and novel DNN models by accessing a few digital and physical images. To the best of our knowledge, this is the first work to formulate the physical attack as a few-shot learning problem. Here, the training task is redefined as the composition of a support set, a query set, and a target DNN model. Under the few-shot setting, we design a novel class-agnostic and model-agnostic meta-learning algorithm to enhance the generalization ability of our method. Extensive experimental results on two benchmark datasets with four challenging experimental settings verify the superior robustness and generalization of our method by comparing to state-of-the-art physical attack methods.

Geometry Projection is a powerful depth estimation method in monocular 3D object detection. It estimates depth dependent on heights, which introduces mathematical priors into the deep model. But projection process also introduces the error amplification problem, in which the error of the estimated height will be amplified and reflected greatly at the output depth. This property leads to uncontrollable depth inferences and also damages the training efficiency. In this paper, we propose a Geometry Uncertainty Projection Network (GUP Net) to tackle the error amplification problem at both inference and training stages. Specifically, a GUP module is proposed to obtains the geometry-guided uncertainty of the inferred depth, which not only provides high reliable confidence for each depth but also benefits depth learning. Furthermore, at the training stage, we propose a Hierarchical Task Learning strategy to reduce the instability caused by error amplification. This learning algorithm monitors the learning situation of each task by a proposed indicator and adaptively assigns the proper loss weights for different tasks according to their pre-tasks situation. Based on that, each task starts learning only when its pre-tasks are learned well, which can significantly improve the stability and efficiency of the training process. Extensive experiments demonstrate the effectiveness of the proposed method. The overall model can infer more reliable object depth than existing methods and outperforms the state-of-the-art image-based monocular 3D detectors by 3.74% and 4.7% AP40of the car and pedestrian categories on the KITTI benchmark. The code and model will be released at https://github.com/SuperMHP/GUPNet.

Weakly supervised object localization (WSOL) aims to localize objects with only image-level labels, which has better scalability and practicability than fully supervised methods in the actual deployment. However, with only image-level labels, learning object classification models tends to activate object parts and ignore the whole object, while expanding object parts into the whole object may deteriorate classification performance. To alleviate this problem, we propose foreground activation maps (FAM), whose aim is to optimize object localization and classification jointly via an object-aware attention module and a part-aware attention module in a unified model, where the two tasks can complement and enhance each other. To the best of our knowledge, this is the first work that can achieve remarkable performance for both tasks by optimizing them jointly via FAM for WSOL. Besides, the designed two modules can effectively highlight foreground objects for localization and discover discriminative parts for classification. Extensive experiments with four backbones on two standard benchmarks demonstrate that our FAM performs favorably against state-of-the-art WSOL methods.

Few-Shot Learning (FSL) aims at classifying samples into new unseen classes with only a handful of labeled samples available. However, most of the existing methods are based on the image-level pooled representation, yet ignore considerable local clues that are transferable across tasks. To address this issue, we propose an end-to-end Task-aware Part Mining Network (TPMN) by integrating an automatic part mining process into the metric-based model for FSL. The proposed TPMN model enjoys several merits. First, we design a meta filter learner to generate task-aware part filters based on the task embedding in a meta-learning way. The task-aware part filters can adapt to any individual task and automatically mine task-related local parts even for an unseen task. Second, an adaptive importance generator is proposed to identify key local parts and assign adaptive importance weights to different parts. To the best of our knowledge, this is the first work to automatically exploit the task-aware local parts in a meta-learning way for FSL. Extensive experimental results on four standard benchmarks demonstrate that the proposed model performs favorably against state-of-the-art FSL methods.

With the explosive growth of video categories, zero-shot learning (ZSL) in video classification has become a promising research direction in pattern analysis and machine learning. Based on some auxiliary information such as word embeddings and attributes, the key to a robust ZSL method is to transfer the learned knowledge from seen classes to unseen classes, which requires relationship modeling between these concepts (e.g., categories and attributes). However, most existing approaches ignore to model the explicit relationships in an end-to-end manner, resulting in low effectiveness of knowledge transfer. To tackle this problem, we reconsider the video ZSL task as a task-driven message passing process to jointly enjoy several merits including alleviated heterogeneity gap, low domain shift, and robust temporal modeling. Specifically, we propose a prototype-sample GNN (PS-GNN) consisting of a prototype branch and a sample branch to directly and adaptively model all the relationships between category-attribute, category-category, and attribute-attribute. The prototype branch aims to learn robust representations of video categories, which takes as input a set of word-embedding vectors corresponding to the concepts. The sample branch is designed to generate features of a video sample by leveraging its object semantics. With the co-adaption and cooperation between both branches, a unified and robust ZSL framework is achieved. Extensive experiments strongly evidence that PS-GNN obtains favorable performance on five popular video benchmarks consistently.

To establish robust semantic correspondence between images covering different objects belonging to the same category, there are three important types of information including inter-image relationship, intra-image relationship and cycle consistency. Most existing methods only exploit one or two types of the above information and cannot make them enhance and complement each other. Different from existing methods, we propose a novel end-to-end Consistency Graph Modeling Network (CGMNet) for semantic correspondence by modeling inter-image relationship, intra-image relationship and cycle consistency jointly in a unified deep model. The proposed CGMNet enjoys several merits. First, to the best of our knowledge, this is the first work to jointly model the three kinds of information in a deep model for semantic correspondence. Second, our model has designed three effective modules including cross-graph module, intra-graph module and cycle consistency module, which can jointly learn more discriminative feature representations robust to local ambiguities and background clutter for semantic correspondence. Extensive experimental results show that our algorithm performs favorably against state-of-the-art methods on four challenging datasets including PF-PASCAL, PF-WILLOW, Caltech-101 and TSS.

Weakly supervised temporal action detection has better scalability and practicability than fully supervised action detection in reality deployment. However, it is difficult to learn a robust model without temporal action boundary annotations. In this paper, we propose an en-to-end Multi-Scale Structure-Aware Network (MSA-Net) for weakly supervised temporal action detection by exploring both the global structure information of a video and the local structure information of actions. The proposed SA-Net enjoys several merits. First, to localize actions with different durations, each video is encoded into feature representations with different temporal scales. Second, based on the multi-scale feature representation, the proposed model has designed two effective structure modeling mechanisms including global structure modeling and local structure modeling, which can effectively learn discriminative structure aware representations for robust and complete action detection. To the best of our knowledge, this is the first work to fully explore the global and local structure information in a unified deep model for weakly supervised action detection. And extensive experimental results on two benchmark datasets demonstrate that the proposed MSA-Net performs favorably against state-of-the-art methods.

Weakly supervised temporal sentence grounding has better scalability and practicability than fully supervised methods in real-world application scenarios. However, most of existing methods cannot model the fine-grained video-text local correspondences well and do not have effective supervision information for correspondence learning, thus yielding unsatisfying performance. To address the above issues, we propose an end-to-end Local Correspondence Network (LCNet) for weakly supervised temporal sentence grounding. The proposed LCNet enjoys several merits. First, we represent video and text features in a hierarchical manner to model the fine-grained video-text correspondences. Second, we design a self-supervised cycle-consistent loss as a learning guidance for video and text matching. To the best of our knowledge, this is the first work to fully explore the fine-grained correspondences between video and text for temporal sentence grounding by using self-supervised learning. Extensive experimental results on two benchmark datasets demonstrate that the proposed LCNet significantly outperforms existing weakly supervised methods.

In this paper, we present a method called density-aware convolutional neural network (DensityCNN) to perform the crowd counting task in various crowded scenes. The key idea of the DensityCNN is to utilize high-level semantic information to provide guidance and constraint when generating density maps. To this end, we implement the DensityCNN by adopting a multi-task CNN structure to jointly learn density-level classification and density map estimation. The density-level classification task learns multi-channel semantic features that are aware of the density distributions of the input image. This task is accomplished via our specially designed group-based convolutional structure in a supervised learning manner. In the density map estimation task, these semantic features are deployed together with high-dimension convolutional features to generate density maps with lower count errors. Extensive experiments on four challenging crowd datasets (ShanghaiTech, UCF_CC_50, UCF-QNCF, and WorldExpo'10) and one vehicle dataset TRANCOS demonstrate the effectiveness of the proposed method.

Person re-identification aims to match person of interest under non-overlapping camera views. Therefore, how to generate a robust and discriminative representation is crucial for person re-identification. Mining local clues from human body parts to describe pedestrians has been extensively studied in existing methods. However, existing methods locate human body parts coarsely and do not consider the relations among different local parts. To address the above problem, we propose a Part-based Structured Representation Learning (PSRL) for better exploiting local clues to improve the person representation. There are two important modules in our architecture: Local Semantic Feature Extraction and Structured Person Representation Learning. The Local Semantic Feature Extraction module is designed to extract local features from human body semantic regions. After obtaining the local features, the Structured Person Representation Learning is proposed to fuse the local features by considering the person structure. To model the underlying person structure, a graph convolutional network is employed to capture the relations of different semantic regions. The generated structured feature encodes underlying person structure information, and local semantic feature can solve the misalignment problem caused by pose variations in feature matching. By combining them together, we can improve the descriptive ability of the generated representation. Extensive evaluations on four standard benchmarks show that our proposed method achieves competitive performance against state-of-the-art methods.

RGB-Infrared (IR) person re-identification is very challenging due to the large cross-modality variations between RGB and IR images. The key solution is to learn aligned features to the bridge RGB and IR modalities. However, due to the lack of correspondence labels between every pair of RGB and IR images, most methods try to alleviate the variations with set-level alignment by reducing the distance between the entire RGB and IR sets. However, this set-level alignment may lead to misalignment of some instances, which limits the performance for RGB-IR Re-ID. Different from existing methods, in this paper, we propose to generate cross-modality paired-images and perform both global set-level and fine-grained instance-level alignments. Our proposed method enjoys several merits. First, our method can perform set-level alignment by disentangling modality-specific and modality-invariant features. Compared with conventional methods, ours can explicitly remove the modality-specific features and the modality variation can be better reduced. Second, given cross-modality unpaired-images of a person, our method can generate cross-modality paired images from exchanged images. With them, we can directly perform instance-level alignment by minimizing distances of every pair of images. Extensive experimental results on two standard benchmarks demonstrate that the proposed model favourably against state-of-the-art methods. Especially, on SYSU-MM01 dataset, our model can achieve a gain of 9.2% and 7.7% in terms of Rank-1 and mAP. Code is available at https://github.com/wangguanan/JSIA-ReID.

Convolutional Neural Network (CNN) based methods generally take crowd counting as a regression task by outputting crowd densities. They learn the mapping between image contents and crowd density distributions. Though having achieved promising results, these data-driven counting networks are prone to overestimate or underestimate people counts of regions with different density patterns, which degrades the whole count accuracy. To overcome this problem, we propose an approach to alleviate the counting performance differences in different regions. Specifically, our approach consists of two networks named Density Attention Network (DANet) and Attention Scaling Network (ASNet). DANet provides ASNet with attention masks related to regions of different density levels. ASNet first generates density maps and scaling factors and then multiplies them by attention masks to output separate attention-based density maps. These density maps are summed to give the final density map. The attention scaling factors help attenuate the estimation errors in different regions. Furthermore, we present a novel Adaptive Pyramid Loss (APLoss) to hierarchically calculate the estimation losses of sub-regions, which alleviates the training bias. Extensive experiments on four challenging datasets (ShanghaiTech Part A, UCF_CC_50, UCF-QNRF, and WorldExpo'10) demonstrate the superiority of the proposed approach.

Image-text matching has received growing interest since it bridges vision and language. The key challenge lies in how to learn correspondence between image and text. Existing works learn coarse correspondence based on object co-occurrence statistics, while failing to learn fine-grained phrase correspondence. In this paper, we present a novel Graph Structured Matching Network (GSMN) to learn fine-grained correspondence. The GSMN explicitly models object, relation and attribute as a structured phrase, which not only allows to learn correspondence of object, relation and attribute separately, but also benefits to learn fine-grained correspondence of structured phrase. This is achieved by node-level matching and structure-level matching. The node-level matching associates each node with its relevant nodes from another modality, where the node can be object, relation or attribute. The associated nodes then jointly infer fine-grained correspondence by fusing neighborhood associations at structure-level matching. Comprehensive experiments show that GSMN outperforms state-of-the-art methods on benchmarks, with relative Recall@1 improvements of nearly 7% and 2% on Flickr30K and MSCOCO, respectively. Code will be released at: https://github.com/CrossmodalGroup/GSMN.

Cross-modality person re-identification (cm-ReID) is a challenging but key technology for intelligent video analysis. Existing works mainly focus on learning modality-shared representation by embedding different modalities into a same feature space, lowering the upper bound of feature distinctiveness. In this paper, we tackle the above limitation by proposing a novel cross-modality shared-specific feature transfer algorithm (termed cm-SSFT) to explore the potential of both the modality-shared information and the modality-specific characteristics to boost the reidentification performance. We model the affinities of different modality samples according to the shared features and then transfer both shared and specific features among and across modalities. We also propose a complementary feature learning strategy including modality adaption, project adversarial learning and reconstruction enhancement to learn discriminative and complementary shared and specific features of each modality, respectively. The entire cmSSFTalgorithm can be trained in an end-to-end manner. We conducted comprehensive experiments to validate the superiority ofthe overall algorithm and the effectiveness ofeach component. The proposed algorithm significantly outperforms state-of-the-arts by 22.5% and 19.3% mAP on the two mainstream benchmark datasets SYSU-MM01 and RegDB, respectively.

The key of image and sentence matching is to accurately measure the visual-semantic similarity between an image and a sentence. However, most existing methods make use of only the intra-modality relationship within each modality or the inter-modality relationship between image regions and sentence words for the cross-modal matching task. Different from them, in this work, we propose a novel MultiModality Cross Attention (MMCA) Network for image and sentence matching by jointly modeling the intra-modality and inter-modality relationships of image regions and sentence words in a unified deep model. In the proposed MMCA, we design a novel cross-attention mechanism, which is able to exploit not only the intra-modality relationship within each modality, but also the inter-modality relationship between image regions and sentence words to complement and enhance each other for image and sentence matching. Extensive experimental results on two standard benchmarks including Flickr30K and MS-COCO demonstrate that the proposed model performs favorably against state-of-the-art image and sentence matching methods.

Generalized Zero-Shot Learning (GZSL) aims at recognizing both seen and unseen classes by constructing correspondence between visual and semantic embedding. However, existing methods have severely suffered from the strong bias problem, where unseen instances in target domain tend to be recognized as seen classes in source domain. To address this issue, we propose an end-to-end Self-supervised Domain-aware Generative Network (SDGN) by integrating self-supervised learning into feature generating model for unbiased GZSL. The proposed SDGN model enjoys several merits. First, we design a cross-domain feature generating module to synthesize samples with high fidelity based on class embeddings, which involves a novel target domain discriminator to preserve the domain consistency. Second, we propose a self-supervised learning module to investigate inter-domain relationships, where a set of anchors are introduced as a bridge between seen and unseen categories. In the shared space, we pull the distribution of target domain away from source domain, and obtain domain-aware features with high discriminative power for both seen and unseen classes. To our best knowledge, this is the first work to introduce self-supervised learning into GZSL as a learning guidance. Extensive experimental results on five standard benchmarks demonstrate that our model performs favorably against state-of-the-art GZSL methods.

Vehicle re-identification aims to identify the same vehicle across different surveillance cameras and plays an important role in public security. Existing approaches mainly focus on exploring informative regions or learning an appropriate distance metric. However, they not only neglect the inherent structured relationship between discriminative regions within an image, but also ignore the extrinsic structured relationship among images. The inherent and extrinsic structured relationships are crucial to learning effective vehicle representation. In this paper, we propose a Structured Graph ATtention network (SGAT) to fully exploit these relationships and allow the message propagation to update the features of graph nodes. SGAT creates two graphs for one probe image. One is an inherent structured graph based on the geometric relationship between the landmarks that can use features of their neighbors to enhance themselves. The other is an extrinsic structured graph guided by the attribute similarity to update image representations. Experimental results on two public vehicle re-identification datasets including VeRi-776 and VehicleID have shown that our proposed method achieves significant improvements over the state-of-the-art methods.

Visual content analysis and understanding attract tremendous attention because of its potentially wide range of applications including human activity analysis, automated photo face tagging, multicamera tracking, crowded counting, and biometric security. With recent progress in end-to-end differentiable learning, the accuracy of algorithms has been significantly improved and even outperforms humans in some tasks. In addition, multimodality methods, targeting on making full use of various visual data sources, are further investigated. These developments contribute to the innovations of two core modules for a typical intelligent vision system, i.e., image and video description and recognition, which are critical for the success of the visual content analysis and understanding in more complex and challenging open world.

Unsupervised person re-identification (Re-ID) has better scalability and practicability than supervised Re-ID in the actual deployment. However, it is difficult to learn a discriminative Re-ID model without annotations. To address the above issue, we propose an end-to-end Self-supervised Agent Learning (SAL) algorithm by exploiting a set of agents as a bridge to reduce domain gaps for unsupervised cross-domain person Re- ID. The proposed SAL model enjoys several merits. First, to the best of our knowledge, this is the first work to exploit selfsupervised learning for unsupervised person Re-ID. Second, our model has designed three effective learning mechanisms including supervised label learning in source domain, similarity consistency learning in target domain, and self-supervised learning in cross domain, which can learn domain-invariant yet discriminative representations through the principled lens of agent learning by reducing domain discrepancy adaptively. Extensive experimental results on three standard benchmarks demonstrate that the proposed SAL performs favorably against state-of-the-art unsupervised person Re-ID methods.

The correlation filters based trackers have achieved an excellent performance for object tracking in recent years. However, most existing methods use only one filter but ignore the information of the previous filters. In this paper, we propose a novel online multi-expert learning algorithm for visual tracking. In our proposed scheme, there are former trackers which retain the previous filters, and those trackers will give their predictions in each frame. The current tracker represents the filter of current frame, and both the current tracker and the former trackers constitute our expert ensemble. We use an adaptive Second-order Quantile strategy to learn the weights of each expert, which can take full advantage of all the experts. To simplify our model and remove some bad experts, we prune our models via a minimum entropy criterion. Finally, we propose a new update strategy to avoid the model corruption problem. Extensive experimental results on both OTB2013 and OTB2015 benchmarks demonstrate that our proposed tracker performs favorably against state-of-the-art methods.

Driven by recent advances in human-centered computing, Facial Expression Recognition (FER) has attracted significant attention in many applications. However, most conventional approaches either perform face frontalization on a non-frontal facial image or learn separate classifier for each pose. Different from existing methods, this paper proposes an end-to-end deep learning model that allows to simultaneous facial image synthesis and pose-invariant facial expression recognition by exploiting shape geometry of the face image. The proposed model is based on generative adversarial network (GAN) and enjoys several merits. First, given an input face and a target pose and expression designated by a set of facial landmarks, an identity-preserving face can be generated through guiding by the target pose and expression. Second, the identity representation is explicitly disentangled from both expression and pose variations through the shape geometry delivered by facial landmarks. Third, our model can automatically generate face images with different expressions and poses in a continuous way to enlarge and enrich the training set for the FER task. Our approach is demonstrated to perform well when compared with state-of-the-art algorithms on both controlled and in-the-wild benchmark datasets including Multi-PIE, BU-3DFE, and SFEW.

Facial expression recognition, face synthesis, and face alignment are three coherently related tasks and can be solved in a joint framework. To achieve this goal, in this paper, we propose a novel end-to-end deep learning model by exploiting the expression code, geometry code and generated data jointly for simultaneous pose-invariant facial expression recognition, face image synthesis, and face alignment. The proposed deep model enjoys several merits. First, to the best of our knowledge, this is the first work to address these three tasks jointly in a unified deep model to complement and enhance each other. Second, the proposed model can effectively disentangle the global and local identity representation from different expression and geometry codes. As a result, it can automatically generate facial images with different expressions under arbitrary geometry codes. Third, these three tasks can further boost their performance for each other via our model. Extensive experimental results on three standard benchmarks demonstrate that the proposed deep model performs favorably against state-of-the-art methods on the three tasks.

Cross-modal hashing (CMH) has been widely used for similarity search in multimedia retrieval applications, thanks to low storage cost and fast query speed. However, preserving the content similarities in finite-length hash codes between different data modalities is still challenging due to the existing heterogeneity gap. To further address the crucial bottleneck, we propose a Multi-Level Correlation Adversarial Hashing (MLCAH) algorithm to integrate the multi-level correlation information into hash codes. The proposed MLCAH model enjoys several merits. First, to the best of our knowledge, it is the early attempt of leveraging the multi-level correlation information for cross-modal hashing retrieval. Second, we propose global and local semantic alignment mechanisms, which can effectively encode multi-level correlation information, including global information, local information, and label information into hash codes. Third, a label-consistency attention mechanism with adversarial training is designed for exploiting the local cross-modality similarity from multi-modality data. Extensive evaluations on four benchmarks demonstrate that the proposed model brings significant improvements over several state-of-the-art cross-modal hashing methods.

Recently, with the ever-growing action categories, zero-shot action recognition (ZSAR) has been achieved by automatically mining the underlying concepts (e.g., actions, attributes) in videos. However, most existing methods only exploit the visual cues of these concepts but ignore external knowledge information for modeling explicit relationships between them. In fact, humans have remarkable ability to transfer knowledge learned from familiar classes to recognize unfamiliar classes. To narrow the knowledge gap between existing methods and humans, we propose an end-to-end ZSAR framework based on a structured knowledge graph, which can jointly model the relationships between action-attribute, action-action, and attribute-attribute. To effectively leverage the knowledge graph, we design a novel Two-Stream Graph Convolutional Network (TS-GCN) consisting of a classifier branch and an instance branch. Specifically, the classifier branch takes the semantic-embedding vectors of all the concepts as input, then generates the classifiers for action categories. The instance branch maps the attribute embeddings and scores of each video instance into an attribute-feature space. Finally, the generated classifiers are evaluated on the attribute features of each video, and a classification loss is adopted for optimizing the whole network. In addition, a self-attention module is utilized to model the temporal information of videos. Extensive experimental results on three realistic action benchmarks Olympic Sports, HMDB51 and UCF101 demonstrate the favorable performance of our proposed framework.

Tracking by siamese networks has achieved favorable performance in recent years. However, most of existing siamese methods do not take full advantage of spatial-temporal target appearance modeling under different contextual situations. In fact, the spatial-temporal information can provide diverse features to enhance the target representation, and the context information is important for online adaption of target localization. To comprehensively leverage the spatial-temporal structure of historical target exemplars and get benefit from the context information, in this work, we present a novel Graph Convolutional Tracking (GCT) method for high-performance visual tracking. Specifically, the GCT jointly incorporates two types of Graph Convolutional Networks (GCNs) into a siamese framework for target appearance modeling. Here, we adopt a spatial-temporal GCN to model the structured representation of historical target exemplars. Furthermore, a context GCN is designed to utilize the context of the current frame to learn adaptive features for target localization. Extensive results on 4 challenging benchmarks show that our GCT method performs favorably against state-of-the-art trackers while running around 50 frames per second.

To bridge source and target domains for domain adaptation, there are three important types of information including data structure, domain label, and class label. Most existing domain adaptation approaches exploit only one or two types of this information and cannot make them complement and enhance each other. Different from existing methods, we propose an end-to-end Graph Convolutional Adversarial Network (GCAN) for unsupervised domain adaptation by jointly modeling data structure, domain label, and class label in a unified deep framework. The proposed GCAN model enjoys several merits. First, to the best of our knowledge, this is the first work to model the three kinds of information jointly in a deep model for unsupervised domain adaptation. Second, the proposed model has designed three effective alignment mechanisms including structure-aware alignment, domain alignment, and class centroid alignment, which can learn domain-invariant and semantic representations effectively to reduce the domain discrepancy for domain adaptation. Extensive experimental results on five standard benchmarks demonstrate that the proposed GCAN algorithm performs favorably against state-of-the-art unsupervised domain adaptation methods.

RGB-Infrared (IR) person re-identification is an important and challenging task due to large cross-modality variations between RGB and IR images. Most conventional approaches aim to bridge the cross-modality gap with feature alignment by feature representation learning. Different from existing methods, in this paper, we propose a novel and end-to-end Alignment Generative Adversarial Network (AlignGAN) for the RGB-IR RE-ID task. The proposed model enjoys several merits. First, it can exploit pixel alignment and feature alignment jointly. To the best of our knowledge, this is the first work to model the two alignment strategies jointly for the RGB-IR RE-ID problem. Second, the proposed model consists of a pixel generator, a feature generator and a joint discriminator. By playing a min-max game among the three components, our model is able to not only alleviate the cross-modality and intra-modality variations, but also learn identity-consistent features. Extensive experimental results on two standard benchmarks demonstrate that the proposed model performs favourably against state-of-the-art methods. Especially, on SYSU-MM01 dataset, our model can achieve an absolute gain of 15.4% and 12.9% in terms of Rank-1 and mAP.

Temporal action localization has recently attracted significant interest in the Computer Vision community. However, despite the great progress, it is hard to identify which aspects of the proposed methods contribute most to the increase in localization performance. To address this issue, we conduct ablative experiments on feature extraction methods, fixed-size feature representation methods and training strategies, and report how each influences the overall performance. Based on our findings, we propose a two-stage detector that outperforms the state of the art in THUMOS14, achieving a mAP@tIoU=0.5 equal to 44.20%.

Semantic segmentation suffers from the fact that densely annotated masks are expensive to obtain. To tackle this problem, we aim at learning to segment by only leveraging scribbles that are much easier to collect for supervision. To fully explore the limited pixel-level annotations from scribbles, we present a novel Boundary Perception Guidance (BPG) approach, which consists of two basic components, i.e., prediction refinement and boundary regression. Specifically, the prediction refinement progressively makes a better segmentation by adopting an iterative upsampling and a semantic feature enhancement strategy. In the boundary regression, we employ class-agnostic edge maps for supervision to effectively guide the segmentation network in localizing the boundaries between different semantic regions, leading to producing finer-grained representation of feature maps for semantic segmentation. The experiment results on the PASCAL VOC 2012 demonstrate the proposed BPG achieves mIoU of 73.2% without fully connected Conditional Random Field (CRF) and 76.0% with CRF, setting up the new state-of-the-art in literature.

Learning semantic correspondence between image and text is significant as it bridges the semantic gap between vision and language. The key challenge is to accurately find and correlate shared semantics in image and text. Most existing methods achieve this goal by representing the shared semantic as a weighted combination of all the fragments (image regions or text words), where fragments relevant to the shared semantic obtain more attention, otherwise less. However, despite relevant ones contribute more to the shared semantic, irrelevant ones will more or less disturb it, and thus will lead to semantic misalignment in the correlation phase. To address this issue, we present a novel Bidirectional Focal Attention Network (BFAN), which not only allows to attend to relevant fragments but also diverts all the attention into these relevant fragments to concentrate on them. The main difference with existing works is they mostly focus on learning attention weight while our BFAN focus on eliminating irrelevant fragments from the shared semantic. The focal attention is achieved by preassigning attention based on inter-modality relation, identifying relevant fragments based on intra-modality relation and reassigning attention. Furthermore, the focal attention is jointly applied in both image-to-text and text-to-image directions, which enables to avoid preference to long text or complex image. Experiments show our simple but effective framework significantly outperforms state-of-the-art, with relative [email protected] gains of 2.2% on both Flicr30K and MSCOCO benchmarks.

The video highlight detection task is to localize key elements (moments of user’s major or special interest) in a video. Most of the existing highlight detection approaches extract features from the video segment as a whole without considering the difference of local features spatially. In spatial extent, not all regions are worth watching because some of them only contain the background of the environment without human or other moving objects, especially when there is lots of clutter in the background. To deal with this issue, we propose a novel region-based model which can automatically localize the key elements in a video without any extra supervised annotations. Specifically, the proposed model produces position-sensitive score maps for local regions in the spatial dimension of the video segment, and then aggregates all position-wise scores with position-pooling operation. The regions with higher response values will be extracted as key elements. Thus more effective features of the video segment are obtained to predict the highlight score. The proposed position-sensitive scheme can be easily integrated into an end-to-end fully convolutional network which aims to update parameters via stochastic gradient descent method in the backward propagation to improve the robustness of the model. Extensive experimental results on the YouTube and SumMe datasets demonstrate that the proposed approach achieves significant improvement over state-of-the-art methods.

In this paper, we propose a multi-task correlation particle filter (MCPF) for robust visual tracking. We first present the multi-task correlation filter (MCF) that takes the interdependencies among different object parts and features into account to learn the correlation filters jointly. Next, the proposed MCPF is introduced to exploit and complement the strength of a MCF and a particle filter. Compared with existing tracking methods based on correlation filters and particle filters, the proposed MCPF enjoys several merits. First, it exploits the interdependencies among different features to derive the correlation filters jointly, and makes the learned filters complement and enhance each other to obtain consistent responses. Second, it handles partial occlusion via a part-based representation, and exploits the intrinsic relationship among local parts via spatial constraints to preserve object structure and learn the correlation filters jointly. Third, it effectively handles large scale variation via a sampling scheme by drawing particles at different scales for target object state estimation. Fourth, it shepherds the sampled particles toward the modes of the target state distribution via the MCF, and effectively covers object states well using fewer particles than conventional particle filters, thereby resulting in robust tracking performance and low computational cost. Extensive experimental results on four challenging benchmark datasets demonstrate that the proposed MCPF tracking algorithm performs favorably against the state-of-the-art methods.

Sparse representations have been applied to visual tracking by finding the best candidate region with minimal reconstruction error based on a set of target templates. However, most existing sparse trackers only consider holistic or local representations and do not make full use of the intrinsic structure among and inside target candidate regions, thereby making them less effective when similar objects appear at close proximity or under occlusion. In this paper, we propose a novel structural sparse representation, which not only exploits the intrinsic relationships among target candidate regions and local patches to learn their representations jointly, but also preserves the spatial structure among the local patches inside each target candidate region. For robust visual tracking, we take outliers resulting from occlusion and noise into account when searching for the best target region. Constructed within a Bayesian filtering framework, we show that the proposed algorithm accommodates most existing sparse trackers with respective merits. The formulated problem can be efficiently solved using an accelerated proximal gradient method that yields a sequence of closed form updates. Qualitative and quantitative evaluations on challenging benchmark datasets demonstrate that the proposed tracking algorithm performs favorably against several state-of-the-art methods.

Most existing trackers are either sampling-based or regression-based methods. Sampling-based methods estimate the target state by sampling many target candidates. Although these methods achieve significant performance, they often suffer from a high computational burden. Regression-based methods often learn a computationally efficient regression function to directly predict the geometric distortion between frames. However, most of these methods require large-scale external training videos and are still not very impressive in terms of accuracy. To make both types of methods enhance and complement each other, in this paper, we propose a joint sampling and regression scheme for visual tracking, which leverages the region proposal network by a novel design. Specifically, our method can jointly exploit discriminative target proposal generation and structural target regression to predict target location in a simple feedforward propagation. We evaluate the proposed method on five challenging benchmarks, and extensive experimental results demonstrate that our method performs favorably compared with state-of-the-art trackers with respect to both accuracy and speed.

Unsupervised domain adaptation aims to transfer domain knowledge from existing well-defined tasks to new ones where labels are unavailable. In the real-world applications, domain discrepancy is usually uncontrollable especially for multi-modality data. Therefore, it is significantly motivated to deal with a multi-modality domain adaptation task. As labels are unavailable in a target domain, how to learn semantic multi-modality representations and successfully adapt the classifier from a source to the target domain remain open challenges in a multi-modality domain adaptation task. To deal with these issues, we propose a multi-modality adversarial network (MMAN), which applies stacked attention to learn semantic multi-modality representations and reduces domain discrepancy via adversarial training. Unlike the previous domain adaptation methods, which cannot make full use of source domain categories information, multi-channel constraint is employed to capture fine-grained categories of knowledge that could enhance the discrimination of target samples and boost target performance on single-modality and multi-modality domain adaptation problems. We apply the proposed MMAN to two applications including cross-domain object recognition and cross-domain social event recognition. The extensive experimental evaluations demonstrate the effectiveness of the proposed model for unsupervised domain adaptation.

Facial expression recognition (FER) is a challenging task due to different expressions under arbitrary poses. Most conventional approaches either perform face frontalization on a non-frontal facial image or learn separate classifiers for each pose. Different from existing methods, in this paper, we propose an end-to-end deep learning model by exploiting different poses and expressions jointly for simultaneous facial image synthesis and pose-invariant facial expression recognition. The proposed model is based on generative adversarial network (GAN) and enjoys several merits. First, the encoder-decoder structure of the generator can learn a generative and discriminative identity representation for face images. Second, the identity representation is explicitly disentangled from both expression and pose variations through the expression and pose codes. Third, our model can automatically generate face images with different expressions under arbitrary poses to enlarge and enrich the training set for FER. Quantitative and qualitative evaluations on both controlled and in-the-wild datasets demonstrate that the proposed algorithm performs favorably against state-of-the-art methods.

Video classification has been achieved by automatically mining the underlying concepts (\eg actions, events) in videos, which plays an essential role in intelligent video analysis. However, most existing algorithms only exploit the visual cues of these concepts but ignore external knowledge information for modeling their relationships during the evolution of videos. In fact, humans have remarkable ability to utilize acquired knowledge to reason about the dynamically changing world. To narrow the knowledge gap between existing methods and humans, we propose an end-to-end video classification framework based on a structured knowledge graph, which can model the dynamic knowledge evolution in videos overtime. Here, we map the concepts of videos to the nodes of the knowledge graph. To effectively leverage the knowledge graph, we adopt a graph convLSTM model to not only identify local knowledge structures in each video shot but also model dynamic patterns of knowledge evolution across these shots. Furthermore, a novel knowledge-based attention model is designed by considering the importance of each video shot and relationships between concepts. We show that by using knowledge graphs, our framework is able to improve the performance of various existing methods. Extensive experimental results on two video classification benchmarks UCF101 and Youtube-8M demonstrate the favorable performance of the proposed framework.

Person re-identification (re-id) aims to match a certain person across multiple non-overlapping cameras. It is a challenging task because the same person's appearance can be very different across camera views due to the presence of large pose variations. To overcome this issue, in this paper, we propose a novel unified person re-id framework by exploiting person poses and identities jointly for simultaneous person image synthesis under arbitrary poses and pose-invariant person re-identification. The framework is composed of a GAN based network and two Feature Extraction Networks (FEN), and enjoys following merits. First, it is a unified generative adversarial model for person image generation and person re-identification. Second, a pose estimator is utilized into the generator as a supervisor in the training process, which can effectively help pose transfer and guide the image generation with any desired pose. As a result, the proposed model can automatically generate a person image under an arbitrary pose. Third, the identity-sensitive representation is explicitly disentangled from pose variations through the person identity and pose embedding. Fourth, the learned re-id model can have better generalizability on a new person re-id dataset by using the synthesized images as auxiliary samples. Extensive experimental results on four standard benchmarks including Market-1501 [69], DukeMTMC-reID [40], CUHK03 [23], and CUHK01 [22] demonstrate that the proposed model can perform favorably against state-of-the-art methods.

Facial expression recognition (FER) is a very challenging problem due to different expressions under arbitrary poses. Most conventional approaches mainly perform FER under laboratory controlled environment. Different from existing methods, in this paper, we formulate the FER in the wild as a domain adaptation problem, and propose a novel auxiliary domain guided Cycle-consistent adversarial Attention Transfer model (CycleAT) for simultaneous facial image synthesis and facial expression recognition in the wild. The proposed model utilizes large-scale unlabeled web facial images as an auxiliary domain to reduce the gap between source domain and target domain based on generative adversarial networks (GAN) embedded with an effective attention transfer module, which enjoys several merits. First, the GAN-based method can automatically generate labeled facial images in the wild through harnessing information from labeled facial images in source domain and unlabeled web facial images in auxiliary domain. Second, the class-discriminative spatial attention maps from the classifier in source domain are leveraged to boost the performance of the classifier in target domain. Third, it can effectively preserve the structural consistency of local pixels and global attributes in the synthesized facial images through pixel cycle-consistency and discriminative loss. Quantitative and qualitative evaluations on two challenging in-the-wild datasets demonstrate that the proposed model performs favorably against state-of-the-art methods.

This paper presents a robust joint discriminative appearance model-based tracking method using online random forests and mid-level feature (superpixels). To achieve superpixel-wise discriminative ability, we propose a joint appearance model that consists of two random forest-based models, i.e., the background-target discriminative model (BTM) and the distractor-target discriminative model (DTM). More specifically, the BTM effectively learns discriminative information between the target object and the background. In contrast, the DTM is used to suppress distracting superpixels, which significantly improves the tracker's robustness and alleviates the drifting problem. A novel online random forest regression algorithm is proposed to build the two models. The BTM and DTM are linearly combined into a joint model to compute a confidence map. Tracking results are estimated using the confidence map, in which the position and scale of the target are estimated orderly. Furthermore, we design a model updating strategy to adapt the appearance changes over time by discarding degraded trees of the BTM and DTM and initializing new trees as replacements. We test the proposed tracking method on two large tracking benchmarks, the CVPR2013 tracking benchmark and VOT2014 tracking challenge. Experimental results show that the tracker runs at real-time speed and achieves favorable tracking performance compared with the state-of-the-art methods. The results also suggest that the DTM improves tracking performance significantly and plays an important role in robust tracking.

Most existing part based tracking methods are part-to-part trackers, which usually have two separated steps including part matching and target localization. Different from existing methods, in this paper, we propose a novel part-totarget (P2T) tracker in a unified fashion by inferring target location from parts directly. To achieve this goal, we propose a novel deep regression model for part to target regression in an end-to-end framework via Convolutional Neural Networks. The proposed model is able to not only exploit part context information to preserve object spatial layout structure, but also learn part reliability to emphasize part importance for robust part to target regression. We evaluate the proposed tracker on 4 challenging benchmark sequences, and extensive experimental results demonstrate that our method performs favorably against state-of-the-art trackers because of the powerful capacity of the proposed deep regression model.

In this paper, we propose a novel correlation particle filter (CPF) for robust visual tracking. Instead of a simple combination of a correlation filter and a particle filter, we exploit and complement the strength of each one. Compared with existing tracking methods based on correlation filters and particle filters, the proposed tracker has four major advantages: 1) it is robust to partial and total occlusions, and can recover from lost tracks by maintaining multiple hypotheses; 2) it can effectively handle large-scale variation via a particle sampling strategy; 3) it can efficiently maintain multiple modes in the posterior density using fewer particles than conventional particle filters, resulting in low computational cost; and 4) it can shepherd the sampled particles toward the modes of the target state distribution using a mixture of correlation filters, resulting in robust tracking performance. Extensive experimental results on challenging benchmark data sets demonstrate that the proposed CPF tracking algorithm performs favorably against the state-of-the-art methods.

The video highlight detection task is to localize key elements (moments of user's major or special interest) in a video. Most of existing highlight detection approaches extract features from the video segment as a whole without considering the difference of local features both temporally and spatially. Due to the complexity of video content, this kind of mixed features will impact the final highlight prediction. In temporal extent, not all frames are worth watching because some of them only contain the background of the environment without human or other moving objects. In spatial extent, it is similar that not all regions in each frame are highlights especially when there are lots of clutters in the background. To solve the above problem, we propose a novel three-dimensional (3-D) (spatial+temporal) attention model that can automatically localize the key elements in a video without any extra supervised annotations. Specifically, the proposed attention model produces attention weights of local regions along both the spatial and temporal dimensions of the video segment. The regions of key elements in the video will be strengthened with large weights. Thus, the more effective feature of the video segment is obtained to predict the highlight score. The proposed 3-D attention scheme can be easily integrated into a conventional end-to-end deep ranking model that aims to learn a deep neural network to compute the highlight score of each video segment. Extensive experimental results on the YouTube and SumMe datasets demonstrate that the proposed approach achieves significant improvement over state-of-the-art methods. With the proposed 3-D attention model, video highlights can be accurately retrieved in spatial and temporal dimensions without human supervision in several domains, such as gymnastics, parkour, skating, skiing, surfing, and dog activities, on the public datasets.

Cross-domain data analysis has been becoming more and more important, and can be effectively adopted for many applications. However, it is difficult to propose a unified cross-domain collaborative learning framework for cross-domain analysis in social multimedia, because cross-domain data have multidomain, multimodal, sparse, and supervised properties. In this paper, we propose a generic cross-domain collaborative learning (CDCL) framework via a discriminative nonparametric Bayesian dictionary learning model for cross-domain data analysis. Compared with existing cross-domain learning methods, our proposed model mainly has four advantages: First, to address the domain discrepancy, we utilize the shared domain priors among multiple domains to make them share a common feature space. Second, to exploit the multimodal property, we use the shared modality priors to model the relationship between different modalities. Third, to deal with the sparse property of media data in one domain, our goal is to learn a shared dictionary to bridge different domains and complement each other. Finally, to make use of the supervised property, we exploit class label information to learn the shared discriminative dictionary, and utilize a latent probability vector to select different dictionary elements for representation of each class. Therefore, the proposed model can investigate the superiorities of different sources to supplement and improve each other effectively. In experiments, we have evaluated our model for two important applications including cross-platform event recognition and cross-network video recommendation. The experimental results have showed the effectiveness of our CDCL model for cross-domain analysis.

In this paper, we aim to automatically identify and track the interesting social event from vast amounts of social media data. However, there are two existing challenges: 1) how to model multimodal social event data over time and visualize the topic evolution and 2) how to alleviate the tracking drift problem to boost social event tracking accuracy. We propose a novel online multimodal multiexpert learning algorithm for social event tracking. Compared with existing methods, the proposed model has several advantages: First, it has a nonparametric online multimodal tracking module, which is able to not only automatically learn the number of topics from data over time, but also exploit the multimodal property of the social event. Second, it adopts a novel multiexpert minimization restoration scheme and allows the tracked model to evolve backwards to undo undesirable model updates, which helps alleviate the model drift problem of social event tracking. Third, it is able to not only effectively track the multimodal social event, but also automatically exploit the topic evolution of the social event for a deep understanding with multimodal topics. To evaluate the proposed model, we collect a real-world dataset for research on social event tracking with multimodality information. We have conducted extensive experiments, and both qualitative and quantitative evaluation results have demonstrated the effectiveness of the proposed model.

Vision-based event analysis is difficult because of the following challenges. The first challenge is intraclass variation. Photos uploaded by users are sparsely sampled visual appearances of an event over time. Thus, each photo may only capture a single object or scene of a specific complex event. The second challenge is interclass confusion. Photos related to different events may contain similar objects or scenes. Third, unusual events are characterized by scarcity, and only a few samples are available for use in learning event patterns. In this paper, by considering the photo timestamp, we propose a structured event modeling (SEM) framework for event analysis that exploits the temporal information of visual features and event classes in a photo sequence. Specifically, the temporal event patterns of the photo sequence and the relationships of different photos are jointly learned using deep neural networks (convolutional neural networks and recurrent neural networks) and a conditional random field. We evaluate the proposed SEM framework in two applications: multiclass event recognition and unusual event detection in photo sequences. The results of extensive experiments performed on a public event recognition dataset and a collected unusual event dataset demonstrate the effectiveness of the proposed method.

Video creation is a challenging and highly profession-al task that generally involves substantial manual efforts. To ease this burden, a better approach is to automatically produce new videos based on clips from the massive amount of existing videos according to arbitrary text. In this paper, we formulate video creation as a problem of retrieving a sequence of videos for a sentence stream. To achieve this goal, we propose a novel multimodal recurrent architecture for automatic video production. Compared with existing methods, the proposed model has three major advantages. First, it is the first completely integrated end-to-end deep learning system for real-world production to the best of our knowledge. We are among the first to address the problem of retrieving a sequence of videos for a sentence stream. Second, it can effectively exploit the correspondence between sentences and video clips through semantic consistency modeling. Third, it can model the visual coherence well by requiring that the produced videos should be organized coherently in terms of visual appearance. We have conducted extensive experiments on two applications, including video retrieval and video composition. The qualitative and quantitative results obtained on two public datasets used in the Large Scale Movie Description Challenge 2016 both demonstrate the effectiveness of the proposed model compared with other state-of-the-art algorithms.

In this paper, we propose a multi-task correlation particle filter (MCPF) for robust visual tracking. We first present the multi-task correlation filter (MCF) that takes the interdependencies among different features into account to learn correlation filters jointly. The proposed MCPF is designed to exploit and complement the strength of a MCF and a particle filter. Compared with existing tracking methods based on correlation filters and particle filters, the proposed tracker has several advantages. First, it can shepherd the sampled particles toward the modes of the target state distribution via the MCF, thereby resulting in robust tracking performance. Second, it can effectively handle large-scale variation via a particle sampling strategy. Third, it can effectively maintain multiple modes in the posterior density using fewer particles than conventional particle filters, thereby lowering the computational cost. Extensive experimental results on three benchmark datasets demonstrate that the proposed MCPF performs favorably against the state-of-the-art methods.

Social event is something that occurs at specific place and time associated with some specific actions, and it consists of many stories over time. With the explosion of Web 2.0 platforms, a popular social event that is happening around us and around the world can spread very fast. As a result, social event analysis becomes more and more important for users to understand the whole evolutionary trend of social event over time. However, it is very challenging to do social event analysis because social event data from different social media sites have multi-modal, multi-domain, and large-scale properties. The goal of our research is to design advanced multimedia techniques to deal with the above issues and establish an effective and robust social event analysis framework for social event representation, detection, tracking and evolution analysis. (1) For social event representation, we propose a novel cross-domain collaborative learning algorithm based on non-parametric Bayesian dictionary learning model. It can make use of the shared domain priors and modality priors to collaboratively learn the data's representations by considering the domain discrepancy and the multi-modal property.(2) For social event detection, we propose a boosted multi-modal supervised Latent Dirichlet Allocation model. It can effectively exploit multi-modality information and utilize boosting weighted sampling strategy for large-scale data processing. (3) For social event tracking, we propose a novel multi-modal event topic model, which can effectively model the correlations between textual and visual modalities, and obtain their topics over time. (4) For social event evolution analysis, we propose a novel multi-modal multi-view topic-opinion mining model to conduct fined-grained topic and opinion analysis for social events from multiple social media sites collaboratively. It can discover multi-modal topics and the corresponding opinions over time to understand the evolutionary processes of social event. Extensive experimental results show that the proposed algorithms perform favorably against state-of-the-art methods for social event analysis.

Personalized video recommender systems play an essential role in bridging users and videos. However, most existing video recommendation methods assume that user profiles (interests) are static. In fact, the static assumption is inadequate to reflect users' dynamic interests as time goes by, especially in the online video recommendation scenarios with dramatic changes of video contents and frequent drift of users' interests over different topics. To overcome the above issue, we propose a dynamic recurrent neural network to model users' dynamic interests over time in a unified framework for personalized video recommendation. Furthermore, to build a much more comprehensive recommendation system, the proposed model is designed to exploit video semantic embedding, user interest modeling, and user relevance mining jointly to model users' preferences. By considering these three factors, the RNN model becomes an interest network which can capture users' high level interests effectively. Extensive experimental results on both single-network and cross-network video recommendation scenarios demonstrate the superior performance of the proposed model compared with other state-of-the-art algorithms.

Multitask learning has shown great potentiality for visual tracking under a particle filter framework. However, the recent multitask trackers, which exploit the similarity between all candidates by imposing group sparsity on the candidate representations, have a limitation in robustness due to the diverse sampling of candidates. To deal with this issue, we propose a discriminative reverse sparse tracker via weighted multitask learning. Our positive and negative templates are retained from the target observations and the background, respectively. Here, the templates are reversely represented via the candidates, and the representation of each positive template is viewed as a single task. Compared with existing multitask trackers, the proposed algorithm has the following advantages. First, we regularize the target representations with the ℓ2,1-norm to exploit the similarity shared by the positive templates, which is reasonable because of the target appearance consistency in the tracking process. Second, the valuable prior relationship between the candidates and the templates is introduced into the representation model by a weighted multitask learning scheme. Third, both target information and background information are integrated to generate discriminative scores for enhancing the proposed tracker. The experimental results on challenging sequences show that the proposed algorithm is effective and performs favorably against 12 state-of-the-art trackers.

An effective representation model, which aims to mine the most meaningful information in the data, plays an important role in visual tracking. Some recent particle-filter-based trackers achieve promising results by introducing the low-rank assumption into the representation model. However, their assumed low-rank structure of candidates limits the robustness when facing severe challenges such as abrupt motion. To avoid the above limitation, we propose a temporal restricted reverse-low-rank learning algorithm for visual tracking with the following advantages: 1) the reverse-low-rank model jointly represents target and background templates via candidates, which exploits the low-rank structure among consecutive target observations and enforces the temporal consistency of target in a global level; 2) the appearance consistency may be broken when target suffers from sudden changes. To overcome this issue, we propose a local constraint via 11,2 mixed-norm, which can not only ensures the local consistency of target appearance, but also tolerates the sudden changes between two adjacent frames; and 3) to alleviate the inference of unreasonable representation values due to outlier candidates, an adaptive weighted scheme is designed to improve the robustness of the tracker. By evaluating on 26 challenge video sequences, the experiments show the effectiveness and favorable performance of the proposed algorithm against 12 state-of-the-art visual trackers.

Most existing tracking methods are direct trackers, which directly exploit foreground or/and background information for object appearance modeling and decide whether an image patch is target object or not. As a result, these trackers cannot perform well when target appearance changes heavily and becomes different from its model. To deal with this issue, we propose a novel relative tracker, which can effectively exploit the relative relationship among image patches from both foreground and background for object appearance modeling. Different from direct trackers, the proposed relative tracker is robust to localize target object by use of the best image patch with the highest relative score to target appearance model. To model relative relationship among large-scale image patch pairs, we propose a novel and effective deep relative learning algorithm via Convolutional Neural Network. We test the proposed approach on challenging sequences involving heavy occlusion, drastic illumination changes, and large pose variations. Experimental results show that our method consistently outperforms state-of-the-art trackers due to the powerful capacity of the proposed deep relative model.

In this paper, we present a part-based sparse tracker in a particle filter framework where both the motion and appearance model are formulated in 3D. The motion model is adaptive and directed according to a simple yet powerful occlusion handling paradigm, which is intrinsically fused in the motion model. Also, since 3D trackers are sensitive to synchronization and registration noise in the RGB and depth streams, we propose automated methods to solve these two issues. Extensive experiments are conducted on a popular RGBD tracking benchmark, which demonstrate that our tracker can achieve superior results, outperforming many other recent and state-of-the-art RGBD trackers.

In this paper, we propose a novel structural correlation filter (SCF) model for robust visual tracking. The proposed SCF model takes part-based tracking strategies into account in a correlation filter tracker, and exploits circular shifts of all parts for their motion modeling to preserve target object structure. Compared with existing correlation filter trackers, our proposed tracker has several advantages: (1) Due to the part strategy, the learned structural correlation filters are less sensitive to partial occlusion, and have computational efficiency and robustness. (2) The learned filters are able to not only distinguish the parts from the background as the traditional correlation filters, but also exploit the intrinsic relationship among local parts via spatial constraints to preserve object structure. (3) The learned correlation filters not only make most parts share similar motion, but also tolerate outlier parts that have different motion. Both qualitative and quantitative evaluations on challenging benchmark image sequences demonstrate that the proposed SCF tracking algorithm performs favorably against several state-of-the-art methods.

Sparse representation has been introduced to visual tracking by finding the best target candidate with minimal reconstruction error within the particle filter framework. However, most sparse representation based trackers have high computational cost, less than promising tracking performance, and limited feature representation. To deal with the above issues, we propose a novel circulant sparse tracker (CST), which exploits circulant target templates. Because of the circulant structure property, CST has the following advantages: (1) It can refine and reduce particles using circular shifts of target templates. (2) The optimization can be efficiently solved entirely in the Fourier domain. (3) High dimensional features can be embedded into CST to significantly improve tracking performance without sacrificing much computation time. Both qualitative and quantitative evaluations on challenging benchmark sequences demonstrate that CST performs better than all other sparse trackers and favorably against state-of-the-art methods.

In this paper, we propose a novel multi-modal multi-view topic-opinion mining (MMTOM) model for social event analysis in multiple collection sources. Compared with existing topic-opinion mining methods, our proposed model has several advantages: (1) The proposed MMTOM can effectively take into account multi-modal and multi-view properties jointly in a unified and principled way for social event modeling. (2) Our model is general and can be applied to many other applications in multimedia, such as opinion mining and sentiment analysis, multi-view association visualization, and topic-opinion mining for movie review. (3) The proposed MMTOM is able to not only discover multi-modal common topics from all collections as well as summarize the similarities and differences of these collections along each specific topic, but also automatically mine multi-view opinions on the learned topics across different collections. (4) Our topic-opinion mining results can be effectively applied to many applications including multi-modal multi-view topic-opinion retrieval and visualization, which achieve much better performance than existing methods. To evaluate the proposed model, we collect a real-world dataset for research on multi-modal multi-view social event analysis, and will release it for academic use. We have conducted extensive experiments, and both qualitative and quantitative evaluation results have demonstrated the effectiveness of the proposed MMTOM.

Vision based event analysis plays a very critical role in automatically organizing user generated photos. As one of the important tasks in event analysis, abnormal event discovery still does not obtain much attentions. It is difficult because only few samples can be used for event pattern learning. In this paper, by considering the photo taken time, we propose a novel one-class structured event modeling (OSEM) where we explore the temporal event patterns in negative photos of the event using the continuous conditional random field (CRF). With the estimated piecewise training of CRF, the proposed OSEM can be efficiently solved using stochastic gradients descent (SGD) in an end-to-end form. The extensive experimental results on a collected abnormal event dataset demonstrate the effectiveness of the proposed OSEM.

In this paper, we formulate particle filter-based object tracking as an exclusive sparse learning problem that exploits contextual information. To achieve this goal, we propose the context-aware exclusive sparse tracker (CEST) to model particle appearances as linear combinations of dictionary templates that are updated dynamically. Learning the representation of each particle is formulated as an exclusive sparse representation problem, where the overall dictionary is composed of multiple group dictionaries that can contain contextual information. With context, CEST is less prone to tracker drift. Interestingly, we show that the popular L1 tracker is a special case of our CEST formulation. The proposed learning problem is efficiently solved using an accelerated proximal gradient method that yields a sequence of closed form updates. To make the tracker much faster, we reduce the number of learning problems to be solved by using the dual problem to quickly and systematically rank and prune particles in each frame. We test our CEST tracker on challenging benchmark sequences that involve heavy occlusion, drastic illumination changes, and large pose variations. Experimental results show that CEST consistently outperforms state-of-the-art trackers.

With the massive growth of social events in Internet , it has become more and more difficult to exactly find and organize the interesting events from massive social media data, which is useful to browse, search, and monitor social events by users or governments . To deal with this problem, we propose a novel multi-modal social event tracking and evolution framework to not only effectively capture multi-modal topics of social events, but also obtain the evolutionary trends of social events and generate effective event summary details over time. To achieve this goal, we propose a novel multi-modal event topic model (mmETM), which can effectively model social media documents, including long text with related images, and learn the correlations between textual and visual modalities to separate the visual-representative topics and non-visual-representative topics. To apply the mmETM model to social event tracking, we adopt an incremental learning strategy denoted as incremental mmETM, which can obtain informative textual and visual topics of social events over time to help understand these events and their evolutionary trends. To evaluate the effectiveness of our proposed algorithm, we collect a real-world dataset to conduct various experiments. Both qualitative and quantitative evaluations demonstrate that the proposed mmETM algorithm performs favorably against several state-of-the-art methods.

Relative attribute (RA) learning aims to learn the ranking function describing the relative strength of the attribute. Most of current learning approaches learn a linear ranking function for each attribute by use of the hand-crafted visual features. Different from the existing study, in this paper, we propose a novel deep relative attributes (DRA) algorithm to learn visual features and the effective nonlinear ranking function to describe the RA of image pairs in a unified framework. Here, visual features and the ranking function are learned jointly, and they can benefit each other. The proposed DRA model is comprised of five convolutional neural layers, five fully connected layers, and a relative loss function which contains the contrastive constraint and the similar constraint corresponding to the ordered image pairs and the unordered image pairs, respectively. To train the DRA model effectively, we make use of the transferred knowledge from the large scale visual recognition on ImageNet [1] to the RA learning task. We evaluate the proposed DRA model on three widely used datasets. Extensive experimental results demonstrate that the proposed DRA model consistently and significantly outperforms the state-of-the-art RA learning methods. On the public OSR, PubFig, and Shoes datasets, compared with the previous RA learning results [2], the average ranking accuracies have been significantly improved by about 8%, 9%, and 14%, respectively.

Content-based video understanding is extremely difficult due to the semantic gap between low-level vision signals and the various semantic concepts (object, action, and scene) in videos. Though feature extraction from videos has achieved significant progress, most of the previous methods rely only on low-level features, such as the appearance and motion features. Recently, visual-feature extraction has been improved significantly with machine-learning algorithms, especially deep learning. However, there is still not enough work focusing on extracting semantic features from videos directly. The goal of this article is to adopt unlabeled videos with the help of text descriptions to learn an embedding function, which can be used to extract more effective semantic features from videos when only a few labeled samples are available for video recognition. To achieve this goal, we propose a novel embedding convolutional neural network (ECNN). We evaluate our algorithm by comparing its performance on three challenging benchmarks with several popular state-of-the-art methods. Extensive experimental results show that the proposed ECNN consistently and significantly outperforms the existing methods.

Sparse representation has been applied to visual tracking by finding the best target candidate with minimal reconstruction error by use of target templates. However, most sparse representation based trackers only consider holistic or local representations and do not make full use of the intrinsic structure among and inside target candidates, thereby making the representation less effective when similar objects appear or under occlusion. In this paper, we propose a novel Structural Sparse Tracking (SST) algorithm, which not only exploits the intrinsic relationship among target candidates and their local patches to learn their sparse representations jointly, but also preserves the spatial layout structure among the local patches inside each target candidate. We show that our SST algorithm accommodates most existing sparse trackers with the respective merits. Both qualitative and quantitative evaluations on challenging benchmark image sequences demonstrate that the proposed SST algorithm performs favorably against several state-of-the-art methods.

Cross-domain data analysis is one of the most important tasks in social multimedia. It has a wide range of real-world applications, including cross-platform event analysis, cross-domain multi-event tracking, cross-domain video recommendation, etc. It is also very challenging because the data have multi-modal and multi-domain properties, and there are no explicit correlations to link different domains. To deal with these issues, we propose a generic Cross-Domain Collaborative Learning (CDCL) framework based on non-parametric Bayesian dictionary learning model for cross-domain data analysis. In the proposed CDCL model, it can make use of the shared domain priors and modality priors to collaboratively learn the data's representations by considering the domain discrepancy and the multi-modal property. As a result, our CDCL model can effectively explore the virtues of different information sources to complement and enhance each other for cross-domain data analysis. To evaluate the proposed model, we apply it for two different applications: cross-platform event recognition and cross-network video recommendation. The extensive experimental evaluations well demonstrate the effectiveness of the proposed algorithm for cross-domain data analysis.

Vision-based sign language recognition has attracted more and more interest from researchers in the computer vision field. In this article, we propose a novel algorithm to model and recognize sign language performed in front of a Microsoft Kinect sensor. Under the assumption that some frames are expected to be both discriminative and representative in a sign language video, we first assign a binary latent variable to each frame in training videos for indicating its discriminative capability, then develop a latent support vector machine model to classify the signs, as well as localize the discriminative and representative frames in each video. In addition, we utilize the depth map together with the color image captured by the Kinect sensor to obtain a more effective and accurate feature to enhance the recognition accuracy. To evaluate our approach, we conducted experiments on both word-level sign language and sentence-level sign language. An American Sign Language dataset including approximately 2,000 word-level sign language phrases and 2,000 sentence-level sign language phrases was collected using the Kinect sensor, and each phrase contains color, depth, and skeleton information. Experiments on our dataset demonstrate the effectiveness of the proposed method for sign language recognition.

In the Web 2.0 era, a huge number of media data, such as text, image/video, and social interaction information, have been generated on the social media sites (e.g., Facebook, Google, Flickr, and YouTube). These media data can be effectively adopted for many applications (e.g., image/video annotation, image/video retrieval, and event classification) in multimedia. However, it is difficult to design an effective feature representation to describe these data because they have multi-modal property (e.g., text, image, video, and audio) and multi-domain property (e.g., Flickr, Google, and YouTube). To deal with these issues, we propose a novel cross-domain feature learning (CDFL) algorithm based on stacked denoising auto-encoders. By introducing the modal correlation constraint and the cross-domain constraint in conventional auto-encoder, our CDFL can maximize the correlations among different modalities and extract domain invariant semantic features simultaneously. To evaluate our CDFL algorithm , we apply it to three important applications: sentiment classification, spam filtering, and event classification. Comprehensive evaluations demonstrate the encouraging performance of the proposed approach.

Vision-based event analysis is extremely difficult due to the various concepts (object, action, and scene) contained in videos. Though visual concept-based event analysis has achieved significant progress, it has two disadvantages: visual concept is defined manually, and has only one corresponding classifier in traditional methods. To deal with these issues, we propose a novel automatic visual concept learning algorithm for social event understanding in videos. First, instead of defining visual concept manually, we propose an effective automatic concept mining algorithm with the help of Wikipedia, N-gram Web services, and Flickr. Then, based on the learned visual concept, we propose a novel boosting concept learning algorithm to iteratively learn multiple classifiers for each concept to enhance its representative discriminability. The extensive experimental evaluations on the collected dataset well demonstrate the effectiveness of the proposed algorithm for social event understanding.

Conventional learning algorithm assumes that the training data and test data share a common distribution. However, this assumption will greatly hinder the practical application of the learned model for cross-domain data analysis in multimedia. To deal with this issue, transfer learning based technology should be adopted. As a typical version of transfer learning, domain adaption has been extensively studied recently due to its theoretical value and practical interest. In this article, we propose a boosted multifeature learning (BMFL) approach to iteratively learn multiple representations within a boosting procedure for unsupervised domain adaption. The proposed BMFL method has a number of properties. (1) It reuses all instances with different weights assigned by the previous boosting iteration and avoids discarding labeled instances as in conventional methods. (2) It models the instance weight distribution effectively by considering the classification error and the domain similarity, which facilitates learning new feature representation to correct the previously misclassified instances. (3) It learns multiple different feature representations to effectively bridge the source and target domains. We evaluate the BMFL by comparing its performance on three applications: image classification, sentiment classification and spam filtering. Extensive experimental results demonstrate that the proposed BMFL algorithm performs favorably against state-of-the-art domain adaption methods.

Part-based visual tracking is advantageous due to its ro-bustness against partial occlusion. However, how to effec-tively exploit the confidence scores of individual parts to construct a robust tracker is still a challenging problem. In this paper, we address this problem by simultaneously matching parts in each of multiple frames, which is realized by a locality-constrained low-rank sparse learning method that establishes multi-frame part correspondences through optimization of partial permutation matrices. The proposed part matching tracker (PMT) has a number of attractive properties. (1) It exploits the spatial-temporal locality-constrained property for robust part matching. (2) It match-es local parts from multiple frames jointly by considering their low-rank and sparse structure information, which can effectively handle part appearance variations due to occlu-sion or noise. (3) The proposed PMT model has the inbuilt mechanism of leveraging multi-mode target templates, so that the dilemma of template updating when encountering occlusion in tracking can be better handled. This contrasts with existing methods that only do part matching between a pair of frames. We evaluate PMT and compare with 10 pop-ular state-of-the-art methods on challenging benchmarks. Experimental results show that PMT consistently outperfor-m these existing trackers. 1.

With the rapidly increasing popularity of Social Media sites (e.g., Flickr, YouTube, and Facebook), it is convenient for users to share their own comments on many social events, which successfully facilitates social event generation, sharing and propagation and results in a large amount of user-contributed media data (e.g., images, videos, and texts) for a wide variety of real-world events of different types and scales. As a consequence, it has become more and more difficult to find exactly the interesting events from massive social media data, which is useful to browse, search and monitor social events by users or governments. To deal with these issues, we propose a novel boosted multi-modal supervised Latent Dirichlet Allocation (BMM-SLDA) for social event classification. Our BMM-SLDA has a number of advantages. (1) It can effectively exploit the multi-modality and the supervised information of social events jointly. (2) It is suitable to large-scale data analysis by utilizing boosting weighted sampling strategy to iteratively select a small subset data to efficiently train the corresponding topic models. (3) It effectively exploits boosting document weight distribution by classification error, and can iteratively learn new topic model to correct the previously misclassified documents. We evaluate our BMM-SLDA on a real-world dataset and show extensive results, which show that our model outperforms state-of-the-art methods.

With the rapidly increasing popularity of social media sites (e.g., Flickr, YouTube, and Facebook), it is convenient for users to share their own comments on many social events, which successfully facilitates social event generation, sharing and propagation and results in a large amount of user-contributed media data (e.g., images, videos, and text) for a wide variety of real-world events of different types and scales. As a consequence, it has become more and more difficult to exactly find the interesting events from massive social media data, which is useful to browse, search and monitor social events by users or governments. To deal with these issues, we propose a novel boosted multimodal supervised Latent Dirichlet Allocation (BMM-SLDA) for social event classification by integrating a supervised topic model, denoted as multi-modal supervised Latent Dirichlet Allocation (mm-SLDA), in the boosting framework. Our proposed BMM-SLDA has a number of advantages. (1) Our mm-SLDA can effectively exploit the multimodality and the multiclass property of social events jointly, and make use of the supervised category label information to classify multiclass social event directly. (2) It is suitable for large-scale data analysis by utilizing boosting weighted sampling strategy to iteratively select a small subset of data to efficiently train the corresponding topic models. (3) It effectively exploits social event structure by the document weight distribution with classification error and can iteratively learn new topic model to correct the previously misclassified event documents. We evaluate our BMM-SLDA on a real world dataset and show extensive experimental results, which demonstrate that our model outperforms state-of-the-art methods.

With the massive growth of events on the Internet, efficient organization and monitoring of events becomes a practical challenge. To deal with this problem, we propose a novel CO-PMHT (CO-Probabilistic Multi-Hypothesis Tracking) algorithm for cross-domain multi-event tracking to obtain their informative summary details and evolutionary trends over time. We collect a large-scale dataset by searching keywords on two domains (Gooogle News and Flickr) and downloading both images and textual content for an event. Given the input data, our algorithm can track multiple events in the two domains collaboratively and boost the tracking performance. Specifically, the bridge between two domains is a semantic posterior probability, that avoids the domain gap. After tracking, we can visualize the whole evolutionary process of the event over time and mine the semantic topics of each event for deep understanding and event prediction. The extensive experimental evaluations on the collected dataset well demonstrate the effectiveness of the proposed algorithm for cross-domain multi-event tracking.

In this paper, we propose a low-rank sparse coding (LRSC) method that exploits local structure information among features in an image for the purpose of image-level classification. LRSC represents densely sampled SIFT descriptors, in a spatial neighborhood, collectively as low-rank, sparse linear combinations of code words. As such, it casts the feature coding problem as a low-rank matrix learning problem, which is different from previous methods that encode features independently. This LRSC has a number of attractive properties. (1) It encourages sparsity in feature codes, locality in codebook construction, and low-rankness for spatial consistency. (2) LRSC encodes local features jointly by considering their low-rank structure information, and is computationally attractive. We evaluate the LRSC by comparing its performance on a set of challenging benchmarks with that of 7 popular coding and other state-of-the-art methods. Our experiments show that by representing local features jointly, LRSC not only outperforms the state-of-the-art in classification accuracy but also improves the time complexity of methods that use a similar sparse linear representation model for feature coding.

In this paper, we propose a novel algorithm to model and recognize sign language with Kinect sensor. We assume that in a sign language video, some frames are expected to be both discriminative and representative. Under this assumption, each frame in training videos is assigned a binary latent variable indicating its discriminative capability. A Latent Support Vector Machine model is then developed to classify the signs, as well as localize the discriminative and representative frames in videos. In addition, we utilize the depth map together with color image captured by Kinect sensor to obtain more effective and accurate feature to enhance the recognition accuracy. To evaluate our approach, we collected an American Sign Language (ASL) dataset which included approximately 2000 phrases, while each phrase was captured by Kinect sensor and hence included color, depth and skeleton information. Experiments on our dataset demonstrate the effectiveness of the proposed method for sign language recognition.

Sign language recognition is a growing research area in the field of computer vision. A challenge within it is to model various signs, varying with time resolution, visual manual appearance, and so on. In this paper, we propose a discriminative exemplar coding (DEC) approach, as well as utilizing Kinect sensor, to model various signs. The proposed DEC method can be summarized as three steps. First, a quantity of class-specific candidate exemplars are learned from sign language videos in each sign category by considering their discrimination. Then, every video of all signs is described as a set of similarities between frames within it and the candidate exemplars. Instead of simply using a heuristic distance measure, the similarities are decided by a set of exemplar-based classifiers through the multiple instance learning, in which a positive (or negative) video is treated as a positive (or negative) bag and those frames similar to the given exemplar in Euclidean space as instances. Finally, we formulate the selection of the most discriminative exemplars into a framework and simultaneously produce a sign video classifier to recognize sign. To evaluate our method, we collect an American sign language dataset, which includes approximately 2000 phrases, while each phrase is captured by Kinect sensor with color, depth, and skeleton information. Experimental results on our dataset demonstrate the feasibility and effectiveness of the proposed approach for sign language recognition.

Automated visual surveillance systems are attracting extensive interest due to public security. In this paper, we attempt to mine semantic context information including object-specific context information and scene-specific context information (learned from object-specific context information) to build an intelligent system with robust object detection, tracking, and classification and abnormal event detection. By means of object-specific context information, a cotrained classifier, which takes advantage of the multiview information of objects and reduces the number of labeling training samples, is learned to classify objects into pedestrians or vehicles with high object classification performance. For each kind of object, we learn its corresponding semantic scene-specific context information: motion pattern, width distribution, paths, and entry/exist points. Based on this information, it is efficient to improve object detection and tracking and abnormal event detection. Experimental results demonstrate the effectiveness of our semantic context features for multiple real-world traffic scenes.

In this paper, we formulate object tracking in a particle filter framework as a multi-task sparse learning problem, which we denote as Multi-Task Tracking (MTT). Since we model particles as linear combinations of dictionary templates that are updated dynamically, learning the representation of each particle is considered a single task in MTT. By employing popular sparsity-inducing ℓp, qmixed norms (p ∈ {2, ∞} and q = 1), we regularize the representation problem to enforce joint sparsity and learn the particle representations together. As compared to previous methods that handle particles independently, our results demonstrate that mining the interdependencies between particles improves tracking performance and overall computational complexity. Interestingly, we show that the popular L1tracker [15] is a special case of our MTT formulation (denoted as the L11tracker) when p = q = 1. The learning problem can be efficiently solved using an Accelerated Proximal Gradient (APG) method that yields a sequence of closed form updates. As such, MTT is computationally attractive. We test our proposed approach on challenging sequences involving heavy occlusion, drastic illumination changes, and large pose variations. Experimental results show that MTT methods consistently outperform state-of-the-art trackers.

Multiple player tracking is one of the main building blocks needed in a sports video analysis system. In an uncalibrated camera setting, robust mutli-object tracking can be very difficult due to a number of reasons including the presence of noise, occlusion, fast camera motion, low-resolution image capture, varying viewpoints and illumination changes. To address the problem of multi-object tracking in sports videos, we go beyond the video frame domain and make use of information in a homography transform domain that is denoted the homography field domain. We propose a novel particle filter based tracking algorithm that uses both object appearance information (e.g. color and shape) in the image domain and cross-domain contextual information in the field domain to improve object tracking. In the field domain, the effect of fast camera motion is significantly alleviated since the underlying homography transform from each frame to the field domain can be accurately estimated. We use contextual trajectory information (intra-trajectory and inter-trajectory context) to further improve the prediction of object states within an particle filter framework. Here, intra-trajectory contextual information is based on history tracking results in the field domain, while inter-trajectory contextual information is extracted from a compiled trajectory dataset based on tracks computed from videos depicting the same sport. Experimental results on real world sports data show that our system is able to effectively and robustly track a variable number of targets regardless of background clutter, camera motion and frequent mutual occlusion between targets.

In this paper, we aim at a practical system, magic closet, for automatic occasion-oriented clothing recommendation. Given a user-input occasion, e.g., wedding, shopping or dating, magic closet intelligently suggests the most suitable clothing from the user's own clothing photo album, or automatically pairs the user-specified reference clothing (upper-body or lower-body) with the most suitable one from online shops.

In this work, we propose a weakly supervised graph propagation method to automatically assign the annotated labels at image level to those contextually derived semantic regions. The graph is constructed with the over-segmented patches of the image collection as nodes. Image-level labels are imposed on the graph as weak supervision information over subgraphs, each of which corresponds to all patches of one image, and the contextual information across different images at patch level are then mined to assist the process of label propagation from images to their descendent regions. The ultimate optimization problem is efficiently solved by Convex Concave Programming (CCCP). Extensive experiments on four benchmark datasets clearly demonstrate the effectiveness of our proposed method for the task of collective image parsing. Two extensions including image annotation and concept map based image retrieval demonstrate the proposed image parsing algorithm can effectively aid other vision tasks.

Learning-based video annotation is essential for video analysis and understanding, and many various approaches have been proposed to avoid the intensive labor costs of purely manual annotation. However, there lacks a generic framework due to several difficulties, such as dependence of domain knowledge, insufficiency of training data, no precise localization and in efficacy for large-scale video dataset. In this paper, we propose a novel approach based on semi-supervised learning by means of information from the Internet for interesting event annotation in videos. Concretely, a Fast Graph-based Semi-Supervised Multiple Instance Learning (FGSSMIL) algorithm, which aims to simultaneously tackle these difficulties in a generic framework for various video domains (e.g., sports, news, and movies), is proposed to jointly explore small-scale expert labeled videos and large-scale unlabeled videos to train the models. The expert labeled videos are obtained from the analysis and alignment of well-structured video related text (e.g., movie scripts, web-casting text, close caption). The unlabeled data are obtained by querying related events from the video search engine (e.g., YouTube, Google) in order to give more distributive information for event modeling. Two critical issues of FGSSMIL are: (1) how to calculate the weight assignment for a graph construction, where the weight of an edge specifies the similarity between two data points. To tackle this problem, we propose a novel Multiple Instance Learning Induced Similarity (MILIS) measure by learning instance sensitive classifiers; (2) how to solve the algorithm efficiently for large-scale dataset through an optimization approach. To address this issue, Concave-Convex Procedure (CCCP) and nonnegative multiplicative updating rule are adopted. We perform the extensive experiments in three popular video domains: movies, sports, and news. The results compared with the state-of-the-arts are promising and demonstrate the effectiveness and efficiency of our proposed approach.

Human action recognition and annotation is an active research topic in computer vision. How to model various actions, varying with time resolution, visual appearance, and others, is a challenging task. In this paper, we propose a boosted exemplar learning (BEL) approach to model various actions in a weakly supervised manner, i.e., only action bag-level labels are provided but action instance level ones are not. The proposed BEL method can be summarized as three steps. First, for each action category, amount of class-specific candidate exemplars are learned through an optimization formulation considering their discrimination and co-occurrence. Second, each action bag is described as a set of similarities between its instances and candidate exemplars. Instead of simply using a heuristic distance measure, the similarities are decided by the exemplar-based classifiers through the multiple instance learning, in which a positive (or negative) video or image set is deemed as a positive (or negative) action bag and those frames similar to the given exemplar in Euclidean Space as action instances. Third, we formulate the selection of the most discriminative exemplars into a boosted feature selection framework and simultaneously obtain an action bag-based detector. Experimental results on two publicly available datasets: the KTH dataset and Weizmann dataset, demonstrate the validity and effectiveness of the proposed approach for action recognition. We also apply BEL to learn representations of actions by using images collected from the Web and use this knowledge to automatically annotate action in YouTube videos. Results are very impressive, which proves that the proposed algorithm is also practical in unconstraint environments.

Event detection is essential for the extensively studied video analysis and understanding area. Although various approaches have been proposed for event detection, there is a lack of a generic event detection framework that can be applied to various video domains (e.g. sports, news, movies, surveillance). In this paper, we present a generic event detection approach based on semi-supervised learning and Internet vision. Concretely, a Graph-based Semi-Supervised Multiple Instance Learning (GSSMIL) algorithm is proposed to jointly explore small-scale expert labeled videos and large-scale unlabeled videos to train the event models to detect video event boundaries. The expert labeled videos are obtained from the analysis and alignment of well-structured video related text (e.g. movie scripts, web-casting text, close caption). The unlabeled data are obtained by querying related events from the video search engine (e.g. YouTube) in order to give more distributive information for event modeling. A critical issue of GSSMIL in constructing a graph is the weight assignment, where the weight of an edge specifies the similarity between two data points. To tackle this problem, we propose a novel Multiple Instance Learning Induced Similarity (MILIS) measure by learning instance sensitive classifiers. We perform the thorough experiments in three popular video domains: movies, sports and news. The results compared with the state-of-the-arts are promising and demonstrate our proposed approach is performance-effective.

Activity analysis is a basic task in video surveillance and has become an active research area. However, due to the diversity of moving objects category and their motion patterns, developing robust semantic scene models for activity analysis remains a challenging problem in traffic scenarios. This paper proposes a novel framework to learn semantic scene models. In this framework, the detected moving objects are first classified as pedestrians or vehicles via a co-trained classifier which takes advantage of the multiview information of objects. As a result, the framework can automatically learn motion patterns respectively for pedestrians and vehicles. Then, a graph is proposed to learn and cluster the motion patterns. To this end, trajectory is parameterized and the image is cut into multiple blocks which are taken as the nodes in the graph. Based on the parameters of trajectories, the primary motion patterns in each node (block) are extracted via Gaussian mixture model (GMM), and supplied to this graph. The graph cut algorithm is finally employed to group the motion patterns together, and trajectories are clustered to learn semantic scene models. Experimental results and applications to real world scenes show the validity of our proposed method.