Camouflage Images Generation (CIG) is an emerging research area that focuses on synthesizing images in which objects are harmoniously blended and exhibit high visual consistency with their surroundings. Existing methods perform CIG by either fusing objects into specific backgrounds or outpainting the surroundings via foreground object-guided diffusion. However, they often fail to obtain natural results because they overlook the logical relationship between camouflaged objects and background environments. To address this issue, we propose CT-CIG, a Controllable Text-guided Camouflage Images Generation method that produces realistic and logically plausible camouflage images. Leveraging Large Visual Language Models (VLM), we design a Camouflage-Revealing Dialogue Mechanism (CRDM) to annotate existing camouflage datasets with high-quality text prompts. Subsequently, the constructed image-prompt pairs are utilized to finetune Stable Diffusion, incorporating a lightweight controller to guide the location and shape of camouflaged objects for enhanced camouflage scene fitness. Moreover, we design a Frequency Interaction Refinement Module (FIRM) to capture high-frequency texture features, facilitating the learning of complex camouflage patterns. Extensive experiments, including CLIPScore evaluation and camouflage effectiveness assessment, demonstrate the semantic alignment of our generated text prompts and CT-CIG's ability to produce photorealistic camouflage images.

Point cloud completion is a fundamental task in 3D vision. A persistent challenge in this field is simultaneously preserving fine-grained details present in the input while ensuring the global structural integrity of the completed shape. While recent works leveraging local symmetry transformations via direct regression have significantly improved the preservation of geometric structure details, these methods suffer from two major limitations: (1) These regression-based methods are prone to overfitting which tend to memorize instant-specific transformations instead of learning a generalizable geometric prior. (2) Their reliance on point-wise transformation regression lead to high sensitivity to input noise, severely degrading their robustness and generalization. To address these challenges, we introduce Simba, a novel framework that reformulates point-wise transformation regression as a distribution learning problem. Our approach integrates symmetry priors with the powerful generative capabilities of diffusion models, avoiding instance-specific memorization while capturing robust geometric structures. Additionally, we introduce a hierarchical Mamba-based architecture to achieve high-fidelity upsampling. Extensive experiments across the PCN, ShapeNet, and KITTI benchmarks validate our method's state-of-the-art (SOTA) performance.

Object Goal Navigation (ObjectNav) requires an agent to locate a target object in unseen environments based on partial visual observations. A central challenge lies in reasoning about where the target is likely to appear under semantic uncertainty, sparse observations, and ambiguous scene cues. Existing approaches often rely on extensive training or static semantic priors, limiting adaptability and interpretability in unfamiliar environments. We introduce PRISM, a preference-guided semantic reasoning and mapping framework that integrates vision-language models (VLMs) into the navigation loop to support efficient semantic object search. PRISM constructs a dynamic Preference-guided Semantic Map (PSM) that aggregates observed semantics, predicted unobserved regions, and language-mediated semantic reasoning, enabling the agent to incrementally refine its belief over likely target locations. Based on the PSM, PRISM further employs Adaptive Preference Exploration Strategies (APES) to adjust exploration behavior using contextual semantic cues and historical feedback under semantic ambiguity. Experiments on the Gibson, Matterport3D (MP3D), and Habitat-Matterport3D (HM3D) datasets show that our method achieves consistent improvements over prior training-free approaches, yielding absolute gains of +2.7% SR on Gibson, +2.8% SR on MP3D, and +3.2% SR on HM3D.

Accurate estimation of reflectance and illumination maps in the Retinex framework remains a significant challenge for low-light image enhancement due to inherent decomposition ambiguity. To address this, we propose DualPrior-Retinex, a novel framework that, inspired by Retinex theory, leverages the practical advantages of the YUV color space for robust enhancement. Our framework introduces a dual-prior architecture that effectively decouples the restoration process. It combines a diffusion-based global prior, responsible for ensuring low-frequency content consistency, with a YUV-based local prior designed to preserve high-frequency structural details. These complementary components are integrated by our Hierarchical Prior Fusion Module (HPFM), which balances perceptual quality with pixel-level fidelity in complex low-light scenarios. Extensive evaluations on multiple benchmarks demonstrate that our method achieves state-of-the-art performance across diverse metrics and visual qualities. Codes will be released at https://github.com/I2-Multimedia-Lab/DP-Retinex.

Effective monitoring of unusual transportation behaviors, such as wrong-way cycling (i.e., riding a bicycle or e-bike against designated traffic flow), is crucial for optimizing law enforcement deployment and traffic planning. However, accurately recording all wrong-way cycling events is both unnecessary and infeasible in resource-constrained environments, as it requires high-resolution cameras for evidence collection and event detection. To address this challenge, we propose WWC-Predictor, a novel method for efficiently estimating the wrong-way cycling ratio, defined as the proportion of wrong-way cycling events relative to the total number of cycling movements over a given time period. The core innovation of our method lies in accurately detecting wrong-way cycling events in sparsely sampled frames using a light-weight detector, then estimating the overall ratio using an autoregressive moving average model. To evaluate the effectiveness of our method, we construct a benchmark dataset consisting of 35 minutes of video sequences with minute-level annotations. Our method achieves an average error rate of a mere 1.475% while consuming only 19.12% GPU time required by conventional tracking methods, validating its effectiveness in estimating the wrong-way cycling ratio. Our source code is publicly available at:https://github.com/VICA-Lab-HKUST-GZ/WWC-Predictor

Domain Adaptive Person Search (DAPS) aims to improve the generalization capability of person search models by training on both labeled source data and unlabeled target data, which is not that practical in real-world applications considering the storage/transmission costs and the privacy of source data. In this paper, we investigate a more practical and efficient person search setting, Source-Free Domain Adaptive Person Search (SFDA-PS), which seeks to generalize an existing source person search model to any unseen domain without requiring source data. Considering the absence of effective annotations in SFDA-PS, we propose a Doubly Contrastive Learning (DCL) method to adapt the target domain knowledge to the source model in a mutual learning and contrastive learning way. Specifically, we employ a mutual learning-based mean-teacher model as our baseline to incorporate target domain knowledge by pursuing prediction consistency between the teacher and student. Then, a Relation-embedded Contrastive (ReC) learning strategy is introduced to the detection head to ensure semantic consistency among proposals related to the same person while maintaining semantic distinction among proposals from different categories or persons. Furthermore, a Memory-aided Constrative (MaC) learning strategy is integrated into the re-identification (Re-ID) head to enhance its discriminative capability on target person embeddings. Extensive experiments on existing state-of-the-art person search models and two widely used benchmarks demonstrate the superiority of the proposed SFDA-PS task, as well as our proposed DCL.

Camouflaged Object Detection (COD) presents a formidable challenge in computer vision due to the inherent difficulty of distinguishing the foreground from the background when objects are camouflaged. Inspired by research on optical imaging that proves the significance of depth of field for enhancing visual clarity and object separation, we introduce depth images into the COD framework to augment RGB data. In this paper, we propose a novel FocusNet that mimics the focusing process in photography. Specifically, we first propose a Dual-Modality Fusion (DMF) module to collaboratively learn deep features from both RGB and depth modalities, framing the whole concealed scene. Then, we propose another Depth Feature Focusing (DFF) module to fully exploit the localization cues in depth images, focusing on the foreground region. Finally, we propose the Hierarchical Feature Decoder (HFD) to generate accurate masks from multi-level features, capturing the camouflaged target. Extensive experiments demonstrate the effectiveness of the proposed FocusNet.

Roadside cameras play a crucial role in extending the perception range of intelligent transportation infrastructures, offering a cost-effective means for long-range 3D object detection. However, under foggy conditions, the visibility degradation severely diminishes the performance of vision-based roadside perception systems, as reduced contrast and distorted monocular cues impair bird’s-eye-view (BEV) depth estimation. To address this critical challenge, we propose BEVFog, a novel end-to-end framework that simultaneously restores fog-degraded semantic features and leverages fog density as an auxiliary depth prior to enhance detection robustness. BEVFog integrates two key components: a lightweight, region-aware defogging module with CNN-predicted, differentiable filters, and a depth adjustment module that uses Fog-Cue Conditional Kernels (FCK) to modulate BEV features based on local scattering, correcting depth distortions and enhancing geometric cues. To facilitate rigorous evaluation, we construct Foggy DAIR-V2X-I, a synthetic benchmark that introduces ten fog levels into the large-scale DAIR-V2X-I roadside dataset using an atmospheric scattering model. Extensive experiments demonstrate that BEVFog narrows the performance gap between heavy-fog and clear-weather conditions from over 10% AP to less than 2% AP for vehicles, while achieving up to 8% AP improvements for pedestrians and cyclists compared to the best dehazing-plus-detection baselines. This work paves the way for safer, all-weather autonomous infrastructure perception, and highlights the importance of integrating low-level restoration with high-level 3D reasoning in adverse conditions.

Although DETR-based methods have achieved considerable success in semi-supervised object detection (SSOD), several challenges remain unresolved: (1) To obtain higher-quality pseudo-labels for training, teacher models often adopt larger-scale architectures, which severely impacts inference speed. In contrast, lightweight detectors may compromise the effectiveness of existing SSOD methods. (2) Bipartite matching utilizing the Hungarian algorithm does not fully leverage potentially valuable pseudo-labels. (3) Current methods alleviate the negative impact of low-quality pseudo-labels through one-to-many assignment, yet this often leads to issues such as duplicate detections. To tackle these challenges, we propose a lightweight end-to-end semi-supervised object detection framework called Efficient Semi-DETR. Specifically, to improve accuracy while reducing inference latency, we introduce a heterogeneous teacher-student framework, which leverages a Collaborative Auxiliary Head to better mine potentially valuable pseudo-labels. To tackle the issue of duplicate detections, we propose Efficient Query Matching to improve training efficiency and enhance the detection of small objects. Notably, Efficient Semi-DETR achieves 44.92 mAP with only 10% of the annotated MS-COCO data, surpassing state-of-the-art methods, while its inference latency is only a quarter of existing methods.

Active learning (AL) is an effective method to balance annotation costs and model performance under resource-constrained circumstances. Most existing AL studies are typically designed for class-balanced datasets. However, the ubiquity of long-tailed distributions in real-world scenarios largely restricts the applicability of those AL methods. To tackle this problem, we propose a new active learning framework, namely long-tailed active learning (LTAL). The LTAL framework divides the long-tailed dataset into constantly evolving in-distribution (ID) and out-of-distribution (OOD) samples, and views the tail samples as OOD samples distinct from the head ones, thus intuitively converting the LTA problem into an iterative OOD detection task. We leverage an energy-based OOD detection approach with a well-designed class-imbalanced energy regularization loss to further extend the energy gap between head and tail classes, encouraging the model to select more unlabeled tail samples with higher free energy values. Experimental results show that despite its conceptual simplicity, the proposed method significantly outperforms competitive baselines.

Goal-directed scanpath prediction aims to predict people’s gaze shift path when searching for objects in a visual scene. Most existing goal-directed scanpath prediction methods cannot generalize to target classes not present during training. Besides, they usually exploit different pre-trained models to extract features for the target prompt and image, resulting in big feature gap and making the subsequent feature matching and fusion very difficult. To solve the above problems, we propose a novel zero-shot goal-directed scanpath prediction model named CLIPGaze. We use CLIP to extract pre-matched features for the target prompt and input image, making the feature fusion easier to receive. Using large model like CLIP can also enhance the whole model’s generalization ability on target classes not present during training. We propose a hierarchical visual-semantic feature fusion module to fuse the target and image features more comprehensively. Furthermore, due to the limited number of classes in goal-directed scanpath dataset, we employ image segmentation as a proxy task to help train the feature fusion module, significantly enhancing our model’s performance in zeroshot setting. Extensive experiments demonstrate the effectiveness of our method on both seen and unseen target classes.

Camouflage images generation refers to concealing a salient object within a specific background by integrating its appearance with the surrounding environment. Previous approaches typically perform it by blending foreground and background images, which often suffer from 1) insufficient or excessive concealment, and 2) lack of effective evaluation metrics for assessing camouflage quality. To address these issues, we propose an Adaptive Camouflage Synthesizing network, named ACS-Net, which aims to generate images that effectively camouflage the object and minimize its salience. Specifically, a Multi-level Feature Aggregation Module (MFFM) is designed to improve the granularity of feature extraction, along with a Dual-branch Attention Enhancement Module (DAEM) for fore-back weight importance measurement to avoid excessive concealment. Subsequently, a Position-aligned Feature Fusion Module (PFFM) is utilized to effectively camouflage objects to an optimal extent guided by structural similarity. Furthermore, we employ detection methods for both camouflaged and salient objects to process our generated images as an unbiased evaluation for camouflage quality. Comprehensive experiments demonstrate that ACS-Net outperforms existing methods and is capable of generating high-quality camouflage images with an appropriate balance of concealment and salience. Code is available at https://github.com/NikoNairre/ACS-Net.

Text-based person search aims to identify a person based on textual descriptions, by simultaneously addressing person detection and cross-modal alignment between text queries and person images. Existing approaches often struggle with conflicts in exploiting proposals across these two sub-tasks. Specifically, cross-modal alignment requires highly precise proposals, while person detection can tolerate a certain degree of proposal inaccuracy but always needs a large number of proposals. In this paper, we propose the Decoupled-Coupled Network (DCN) to tackle the above conflicts. We first attempt to resolve the above conflicts by proposing a Decoupled Proposal Selection (DPS) strategy, inspired by the divide-and-conquer principle. DPS adaptively selects the most suitable proposals for each sub-task, ensuring their distinct requirements are adequately met. We further present a Coupled Cascade Refinement (CCR) module to jointly optimize both sub-tasks in a multi-stage manner, progressively improving detection accuracy and fostering cross-modal alignment between text and person image. In addition, we introduce two types of objective functions to optimize the inherently multi-positive contrastive learning challenge. Extensive experiments conducted on two benchmarks demonstrate the effectiveness and superiority of our DCN over existing competitors.

We propose the Multi-modal Untrimmed Video Retrieval task, along with a new benchmark (MUVR) to advance video retrieval for long-video platforms. MUVR aims to retrieve untrimmed videos containing relevant segments using multi-modal queries. It has the following features: 1) Practical retrieval paradigm: MUVR supports video-centric multi-modal queries, expressing fine-grained retrieval needs through long text descriptions, video tag prompts, and mask prompts. It adopts a one-to-many retrieval paradigm and focuses on untrimmed videos, tailored for long-video platform applications. 2) Multi-level visual correspondence: To cover common video categories (e.g., news, travel, dance) and precisely define retrieval matching criteria, we construct multi-level visual correspondence based on core video content (e.g., news events, travel locations, dance moves) which users are interested in and want to retrieve. It covers six levels: copy, event, scene, instance, action, and others. 3) Comprehensive evaluation criteria: We develop 3 versions of MUVR (i.e., Base, Filter, QA). MUVR-Base/Filter evaluates retrieval models, while MUVR-QA assesses MLLMs in a question-answering format. We also propose a Reranking Score to evaluate the reranking ability of MLLMs. MUVR consists of 53K untrimmed videos from the video platform Bilibili, with 1,050 multi-modal queries and 84K matches. Extensive evaluations of 3 state-of-the-art video retrieval models, 6 image-based VLMs, and 10 MLLMs are conducted. MUVR reveals the limitations of retrieval methods in processing untrimmed videos and multi-modal queries, as well as MLLMs in multi-video understanding and reranking. Our code and benchmark is available at https://github.com/debby-0527/MUVR.

Video Anomaly Detection (VAD) aims to locate unusual activities or behaviors within videos. Recently, offline VAD has garnered substantial research attention, which has been invigorated by the progress in large language models (LLMs) and vision-language models (VLMs), offering the potential for a more nuanced understanding of anomalies. However, online VAD has seldom received attention due to real-time constraints and computational intensity. In this paper, we introduce a novel **M**emory-based online scoring queue scheme for **T**raining-free VAD (MoniTor), to address the inherent complexities in online VAD. Specifically, MoniTor applies a streaming input to VLMs, leveraging the capabilities of pre-trained large-scale models. To capture temporal dependencies more effectively, we incorporate a novel prediction mechanism inspired by Long Short-Term Memory (LSTM) networks. This ensures the model can effectively model past states and leverage previous predictions to identify anomalous behaviors. Thereby, it better understands the current frame. Moreover, we design a scoring queue and an anomaly prior to dynamically store recent scores and cover all anomalies in the monitoring scenario, providing guidance for LLMs to distinguish between normal and abnormal behaviors over time. We evaluate MoniTor on two large datasets (i.e., UCF-Crime and XD-Violence) containing various surveillance and real-world scenarios. The results demonstrate that MoniTor outperforms state-of-the-art methods and is competitive with weakly supervised methods without training. Code is available at https://github.com/YsTvT/MoniTor.

Human-annotated attributes serve as effective semantic label embeddings for zero-shot learning (ZSL); however, their annotation is labor-intensive and difficult to scale. Recent studies have explored weakly supervised semantic label embeddings to reduce human effort, but these methods often fail to capture visual similarity and underperform compared to human-annotated semantics. In this work, we propose a minimally supervised yet effective approach: GPT- and CLIP-powered attributes (GCAtt). Specifically, we introduce a three-step interaction process with ChatGPT-comprising preliminary design, hierarchical refinement, and specific value determination-to generate attributes that are both category-shared and discriminative for classification. Additionally, we develop a method that encodes attributes and their values as potential text pairings, leveraging CLIP's retrieval capabilities for annotation. Experimental results on four widely used benchmarks demonstrate that GCAtt consistently outperforms human-annotated semantics. Code and data are available at https://github.com/RowenaHe/GCAtt.

Person search is a challenging task that involves detecting and retrieving individuals from a large set of un-cropped scene images. Existing person search models are mostly trained and deployed in the same-origin scenarios. However, collecting and annotating training samples for each scene is difficult due to the limitation of resources and labor cost. Moreover, large-scale intra-domain data for training are generally not legally available for common developers, due to the regulation of privacy and public security. Leveraging easily accessible large-scale User Generated Video Contents (i.e. UGC videos) to train person search models can fit the real-world distribution, but still suffering a performance gap from the domain difference to surveillance scenes. In this work, we explore enhancing the out-of-domain generalization capabilities of person search models, and propose a generalizable framework on both feature-level and data-level generalization to facilitate downstream tasks in arbitrary scenarios. Specifically, we focus on learning domain-invariant representations for both detection and ReID by introducing a multi-task prototype-based domain-specific batch normalization, and a channel-wise ID-relevant feature decorrelation strategy. We also identify and address typical sources of noise in UGC training frames, including inaccurate bounding boxes, the omission of identity labels, and the absence of cross-camera data. Our framework achieves promising performance on two challenging person search benchmarks without using any human annotation or samples from the target domain. The code is available athttps://github.com/caposerenity/GPS.

Recently, salient object detection (SOD) methods have achieved impressive performance. However, salient regions predicted by existing methods usually contain unsaturated regions and shadows, which limits the model for reliable fine-grained predictions. To address this, we introduce the uncertainty guidance learning approach to SOD, intended to enhance the model's perception of uncertain regions. Specifically, we design a novel Uncertainty Guided Refinement Attention Network (UGRAN), which incorporates three important components, i.e., the Multilevel Interaction Attention (MIA) module, the Scale Spatial-Consistent Attention (SSCA) module, and the Uncertainty Refinement Attention (URA) module. Unlike conventional methods dedicated to enhancing features, the proposed MIA facilitates the interaction and perception of multilevel features, leveraging the complementary characteristics among multilevel features. Then, through the proposed SSCA, the salient information across diverse scales within the aggregated features can be integrated more comprehensively and integrally. In the subsequent steps, we utilize the uncertainty map generated from the saliency prediction map to enhance the model's perception capability of uncertain regions, generating a highly-saturated fine-grained saliency prediction map. Additionally, we devise an adaptive dynamic partition (ADP) mechanism to minimize the computational overhead of the URA module and improve the utilization of uncertainty guidance. Experiments on seven benchmark datasets demonstrate the superiority of the proposed UGRAN over the state-of-the-art methodologies. Codes will be released at https://github.com/I2-Multimedia-Lab/UGRAN.

Person search is a challenging task in computer vision and multimedia understanding, which aims at localizing and identifying target individuals in realistic scenes. State-of-the-art models achieve remarkable success but suffer from overloaded computation and inefficient inference, making them impractical in most real-world applications. A promising approach to tackle this dilemma is to compress person search models with knowledge distillation (KD). Previous KD-based person search methods typically distill the knowledge from the re-identification (re-id) branch, completely overlooking the useful knowledge from the detection branch. In addition, we elucidate that the imbalance between person and background regions in feature maps has a negative impact on the distillation process. To this end, we propose a novel KD-based approach, namely Disaggregation Distillation for Person Search (DDPS), which disaggregates the distillation process and feature maps, respectively. Firstly, the distillation process is disaggregated into two task-oriented sub-processes,i.e., detection distillation and re-id distillation, to help the student learn both accurate localization capability and discriminative person embeddings. Secondly, we disaggregate each feature map into person and background regions, and distill these two regions independently to alleviate the imbalance problem. More concretely, three types of distillation modules,i.e., logit distillation (LD), correlation distillation (CD), and disaggregation feature distillation (DFD), are particularly designed to transfer comprehensive information from the teacher to the student. Note that such a simple yet effective distillation scheme can be readily applied to both homogeneous and heterogeneous teacher-student combinations. We conduct extensive experiments on two person search benchmarks, where the results demonstrate that, surprisingly, our DDPS enables the student model to surpass the performance of the corresponding teacher model, even achieving comparable results with general person search models.

Group activity recognition (GAR) is a challenging task in computer vision, which needs to comprehensively model the spatiotemporal relations among actors. However, most previous methods tend to only model unitary actor relations and directly aggregate actor features to form group representation at a single scale. To address these issues, we propose a novel GAR approach termed multi-scale cross-distance transformer (MSCD-Former), capable of capturing diverse actor relation contexts in multiple spatiotemporal scales. A cross-distance attentive block (CDA-Block) is designed to decompose the actor relations into local and distant ones, diversifying the relation features in rearranged groups. The multi-scale group descriptors are then enhanced by deploying stacked CDA-Blocks to cascaded stages and tightening the sampling scales accordingly. Moreover, we introduce a multi-scale reconstructive learning measure (MSR-Learning) between adjacent scales of CDA-Blocks. Via the reconstruction of actor relational features from lower scales to upper scales, MSR-Learning can enforce semantic consistency in multiple spatiotemporal scales. Consequently, our MSCD-Former boosts GAR by fusing such discriminative relation features of different scales. We extensively evaluate the proposed approach on the VolleyTactic, Volleyball, Collective Activity, NBA, and JRDB-PAR datasets, and the experimental results demonstrate its superiority.

Image Quality Assessment (IQA) has long been a research hotspot in the field of image processing, especially No-Reference Image Quality Assessment (NR-IQA). Due to the powerful feature extraction ability, existing Convolution Neural Network (CNN) and Transformers based NR-IQA methods have achieved considerable progress. However, they still exhibit limited capability when facing unknown authentic distortion datasets. To further improve NR-IQA performance, in this paper, a novel supervised contrastive learning (SCL) and Transformer-based NR-IQA model SaTQA is proposed. We first train a model on a large-scale synthetic dataset by SCL (no image subjective score is required) to extract degradation features of images with various distortion types and levels. To further extract distortion information from images, we propose a backbone network incorporating the Multi-Stream Block (MSB) by combining the CNN inductive bias and Transformer long-term dependence modeling capability. Finally, we propose the Patch Attention Block (PAB) to obtain the final distorted image quality score by fusing the degradation features learned from contrastive learning with the perceptual distortion information extracted by the backbone network. Experimental results on six standard IQA datasets show that SaTQA outperforms the state-of-the-art methods for both synthetic and authentic datasets. Code is available at https://github.com/I2-Multimedia-Lab/SaTQA.

Existing Blind image Super-Resolution (BSR) methods focus on estimating either kernel or degradation infor-mation, but have long overlooked the essential content details. In this paper, we propose a novel BSR approach, Content-aware Degradation-driven Transformer (CDFormer), to capture both degradation and content rep-resentations. However, low-resolution images cannot pro-vide enough content details, and thus we introduce a diffusion-based module CD Former dif f to first learn Con-tent Degradation Prior (CDP) in both low- and high-resolution images, and then approximate the real distribution given only low-resolution information. Moreover, we apply an adaptive SR network CDFormersR that effectively utilizes CDP to refine features. Compared to previous diffusion-based SR methods, we treat the diffusion model as an estimator that can overcome the limitations of expensive sampling time and excessive diversity. Experiments show that CDFormer can outperform existing methods, establishing a new state-of-the-art performance on various bench-marks under blind settings. Codes and models will be avail-able at https://github.com/I2-Multimedia-Lab/CDFormer.

Style transfer aims to render an image with the artistic features of a style image, while maintaining the origi-nal structure. Various methods have been put forward for this task, but some challenges still exist. For instance, it is difficult for CNN-based methods to handle global information and long-range dependencies between input images, for which transformer-based methods have been proposed. Although transformers can better model the relationship between content and style images, they require high-cost hard-ware and time-consuming inference. To address these is-sues, we design a novel transformer model that includes only the encoder, thus significantly reducing the computational cost. In addition, we also find that existing style transfer methods may lead to images under-stylied or missing content. In order to achieve better stylization, we de-sign a content feature extractor and a style feature extrac-tor, based on which pure content and style images can be fed to the transformer. Finally, we propose a novel network termed Puff-Net, i.e., pure content and style feature fusion network. Through qualitative and quantitative experiments, we demonstrate the advantages of our model compared to state-of-the-art ones in the literature. The code is available at https://github.com/ZszYmy9/Puff-Net.

Infrared (IR) images can be seen as complementary to visible light (RGB) images, as they can capture accurate targets in low-visibility conditions. However, camouflaged object detection (COD) based on RGB and IR images is expensive. To this end, we propose to exploit a style transfer-based data augmentation method to generate pseudo-IR images by absorbing the style information of IR images into RGB images, and to perform COD based on RGB and the pseudo-IR images. For RGB and IR-based COD, we propose a novel Edge-guided Uncertainty-aware Fusion Network (EUFNet), to make better use of the complementarity between the two kinds of images. Specifically, an uncertainty-aware fusion module is first proposed to aggregate RGB and IR features by estimating their uncertainties. Then, an edge enhancement module is proposed to extract and enhance the edge information in multiple stages. Lastly, a hierarchical integration module is designed to integrate RGB and IR features with edge cues. Extensive experiments demonstrate the effectiveness of the generated pseudo-IR images as well as the proposed EUFNet. The code is available at https://github.com/csdahunzi/COD.

Person search (PS) aims to simultaneously localize and identify a target person from natural, uncropped images. Existing PS datasets and research works are mostly based on individual images, exhibiting limited practicability in real-world surveillance scenarios. We contend that videos, compared to static images, offer additional temporal information, making searching for the trajectory of the target person from videos more realistic and accurate. In this paper, we propose a new practical and realistic task, namely person search in videos, and a new evaluation metric specifically tailored for it. To fulfill this, we introduce a new PS dataset, namely MTAPS, based on an existing large-scale simulated video dataset. MTA-PS is the first cross-camera PS dataset in virtual videos, consisting of 6 cameras, 60 videos, 1.8K identities, 295.2K frames, 7.3M bounding boxes, and more than 20 minutes per camera, which is challenging and comprehensive, and meanwhile avoids privacy issues. To validate the effectiveness of PS in videos and make full use of the temporal information on our dataset, we also propose a novel framework by seamlessly integrating the three sub-tasks of person detection, tracking, and re-identification. Extensive experiments demonstrate that our method performs favorably over existing counterparts on the newly-introduced MTA-PS dataset. Codes and datasets are available at https://github.com/mtmyyy/MTA-PS.

Generative model-based zero-shot learning (ZSL) approaches usually transform ZSL to supervised learning by generating full visual features for unseen classes, which may contain redundant or noisy information harmful for ZSL classification. In this work, we propose a novel generative framework that fully exploits the visual features by disentangling them into three distinct components, i.e., semantically correlative (SC), visually discriminative (VD), and residual features, respectively. In particular, we propose an encoder-decoder disentanglement framework to learn SC features that align closely with their corresponding semantic label embeddings, as well as VD features that contribute significantly to visual classification accuracy. Additionally, we introduce a mutual information-based loss to disperse the distributions between the discriminative (SC and VD) features and the residual ones, thereby eliminating the redundant information that may hinder the final ZSL classification. Our proposed method is extensively evaluated on four popular ZSL benchmarks, where the experimental results demonstrate its superiority over existing counterparts. Code is available at https://github.com/RowenaHe/VFD-ZSL.

With the rise of new 3D representations like NeRF and 3D Gaussian splatting, creating realistic 3D scenes is easier than ever before. However, the incompatibility of these 3D representations with existing editing software has also introduced unprecedented challenges to 3D editing tasks. Although recent advances in text-to-image generative models have made some progress in 3D editing, these methods either lack precision or require users to manually specify the editing areas in 3D space, complicating the editing process. To overcome these issues, we propose Edit3D, an innovative 3D editing method designed to enhance editing quality. Specifically, we propose a multi-turn editing framework and introduce an attention-driven open-set segmentation (ADSS) technique within this framework. ADSS allows for more precise segmentation of parts, which enhances the editing precision and minimizes interference with pixels in areas that are not being edited. Additionally, we propose a fine-tuning phase, intended to further improve the overall editing quality without compromising the training efficiency. Experiments demonstrate that Edit3D effectively adjusts 3D scenes based on textual instructions. Through continuous and multiple turns of editing, it achieves more intricate combinations, enhancing the diversity of 3D editing effects. Code is available at https://github.com/PeterouZh/Edit3D.

Text-based person search (TBPS) aims to search for the target person in the full image through textual descriptions. The key to addressing this task is to effectively perform cross-modality alignment between text and images. In this paper, we propose a novel TBPS framework, named Memory-Aided Coarse-to-fine Alignment (MACA), to learn an accurate and reliable alignment between the two modalities. Firstly, we introduce a proposal-based alignment module, which performs contrastive learning to accurately align the textual modality with different pedestrian proposals at a coarse-grained level. Secondly, for the fine-grained alignment, we propose an attribute-based alignment module to mitigate unreliable features by aligning text-wise details with image-wise global features. Moreover, we introduce an intuitive memory bank strategy to supplement useful negative samples for more effective contrastive learning, improving the convergence and generalization ability of the model based on the learned discriminative features. Extensive experiments on CUHK-SYSU-TBPS and PRW-TBPS demonstrate the superiority of MACA over state-of-the-art approaches. The code is available at https://github.com/suliangxu/MACA.

Group activity analysis has attracted remarkable attention recently due to the widespread applications in security, entertainment and military. This article targets at learning group activity representations with self-supervision, which differs from the majorities relying heavily on manually annotated labels. Moreover, existing Self-Supervised Learning (SSL) methods for videos are sub-optimal to generate such representations because of the complex context dynamics in group activities. In this article, an end-to-end framework termed Contextualized Relation Predictive Model (Con-RPM) is proposed for self-supervised group activity representation learning with predictive coding. It involves the Serial-Parallel Transformer Encoder (SPTrans-Encoder) to model the context of spatial interactions and temporal variations, and the Hybrid Context Transformer Decoder (HConTrans-Decoder) to predict the future spatio-temporal relations guided by holistic scene context. Additionally, to improve the discriminability and consistency of prediction, we introduce a united loss integrating group-wise and person-wise contrastive losses in frame-level as well as the adversarial loss in global sequence-level. Consequently, our Con-RPM learns robust group representations via describing temporal evolutions of individual relationships and scene semantics explicitly. Extensive experimental results on downstream tasks indicate the effectiveness and generalization of our model in self-supervised learning, and present state-of-the-art performance on the Volleyball, Collective Activity, VolleyTactic, and Choi's New datasets.

Active learning (AL) aims to find a better trade-off between labeling costs and model performance by consciously selecting more informative samples to label. Recently, adversarial approaches have emerged as effective solutions. Most of them leverage generative adversarial networks to align feature distributions of labeled and unlabeled data, upon which discriminators are trained to better distinguish between them. However, these methods fail to consider the relationship between unlabeled samples and decision boundaries, and their training processes are often complex and unstable. To this end, this paper proposes a novel adversarial AL method, namely multi-classifier adversarial optimization for active learning (MAOAL). MAOAL employs task-specific decision boundaries for data alignment while selecting the most informative samples to label. To fulfill this, we introduce a novel classifier class confusion (C3) metric, which represents the classifier discrepancy as the inter-class correlation of classifier outputs. Without any additional hyper-parameters, the C3 metric further reduces the negative impacts of ambiguous samples in the process of distribution alignment and sample selection. More concretely, the network is trained adversarially by adding two auxiliary classifiers, reducing the distribution bias of labeled and unlabeled samples by minimizing the C3 loss between classifiers, while learning tighter decision boundaries and highlighting hard samples by maximizing the C3 loss. Finally, the unlabeled samples with the highest C3 loss are selected to label. Extensive experiments demonstrate the superiority of our approach over state-of-the-art AL methods in terms of image classification and object detection.

Co-salient object detection (CoSOD) aims at detecting common salient objects within a group of relevant source images. Most of the latest works employ the attention mechanism for finding common objects. To achieve accurate CoSOD results with high-quality maps and high efficiency, we propose a novel Memory-aided Contrastive Consensus Learning (MCCL) framework, which is capable of effectively detecting co-salient objects in real time (∼150 fps). To learn better group consensus, we propose the Group Consensus Aggregation Module (GCAM) to abstract the common features of each image group; meanwhile, to make the consensus representation more discriminative, we introduce the Memory-based Contrastive Module (MCM), which saves and updates the consensus of images from different groups in a queue of memories. Finally, to improve the quality and integrity of the predicted maps, we develop an Adversarial Integrity Learning (AIL) strategy to make the segmented regions more likely composed of complete objects with less surrounding noise. Extensive experiments on all the latest CoSOD benchmarks demonstrate that our lite MCCL outperforms 13 cutting-edge models, achieving the new state of the art (∼5.9% and ∼6.2% improvement in S-measure on CoSOD3k and CoSal2015, respectively). Our source codes, saliency maps, and online demos are publicly available at https://github.com/ZhengPeng7/MCCL.

Vision transformers have recently shown strong global context modeling capabilities in camouflaged object detection. However, they suffer from two major limitations: less effective locality modeling and insufficient feature aggregation in decoders, which are not conducive to camou-flaged object detection that explores subtle cues from indistinguishable backgrounds. To address these issues, in this paper, we propose a novel transformer-based Feature Shrinkage Pyramid Network (FSPNet), which aims to hierarchically decode locality-enhanced neighboring transformer features through progressive shrinking for camou-flaged object detection. Specifically, we propose a non-local token enhancement module (NL-TEM) that employs the non-local mechanism to interact neighboring tokens and explore graph-based high-order relations within tokens to enhance local representations of transformers. Moreover, we design a feature shrinkage decoder (FSD) with adjacent interaction modules (AIM), which progressively aggregates adjacent transformer features through a layer-by-layer shrinkage pyramid to accumulate imperceptible but effective cues as much as possible for object information decoding. Extensive quantitative and qualitative experiments demonstrate that the proposed model significantly outperforms the existing 24 competitors on three challenging COD benchmark datasets under six widely-used evaluation metrics. Our code is publicly available at https://github.com/ZhouHuang23/FSPNet.

Recently, open-vocabulary learning has emerged to accomplish segmentation for arbitrary categories of text-based descriptions, which popularizes the segmentation system to more general-purpose application scenarios. However, existing methods devote to designing specialized architectures or parameters for specific segmentation tasks. These customized design paradigms lead to fragmentation between various segmentation tasks, thus hindering the uniformity of segmentation models. Hence in this paper, we propose FreeSeg, a generic framework to accomplish Unified, Universal and Open-Vocabulary Image Segmentation. FreeSeg optimizes an all-in-one network via one-shot training and employs the same architecture and parameters to handle diverse segmentation tasks seamlessly in the inference procedure. Additionally, adaptive prompt learning facilitates the unified model to capture task-aware and category-sensitive concepts, improving model robustness in multi-task and varied scenarios. Extensive experimental results demonstrate that FreeSeg establishes new state-of-the-art results in performance and generalization on three segmentation tasks, which outperforms the best task-specific architectures by a large margin: 5.5% mIoU on semantic segmentation, 17.6% mAP on instance segmentation, 20.1% PQ on panoptic segmentation for the unseen class on COCO. Project page: https://FreeSeg.github.io.

Person search (PS) aims to jointly localize and identify a query person from natural, uncropped images. Existing works unintentionally adopt pedestrians (with similar poses and unchanging clothing) as the query and restrict the application scenarios in surveillance. This is due to that most PS datasets are collected from surveillance cameras with a limited diversity of views, scenes, appearances, etc. In this paper, we study a more general and realistic task in the wild, where we aim to search target persons with a much higher degree of diversity. To this end, we introduce a new PS dataset, namely MovieNet-PS, based on an existing large-scale movie dataset. MovieNet-PS is currently the largest and most diverse PS dataset, consisting of 160K images (100K for training), 274K bounding boxes, and 3K identities. It stands out from existing counterparts from two levels of diversities, i.e., scene-level and identity-level, with 92,043 scenes and significant variations in poses, clothing, scales, etc. for the same identity. To validate the rich context information on our dataset and make full use of it, we propose a novel global-local context network, which exploits scene and group context to boost the search performance. Extensive experiments demonstrate that MovieNet-PS is more challenging and comprehensive than existing datasets, and our approach further pushes the state of the art by a large margin (relatively 34% in mAP) on this dataset. Codes, models, and the dataset are available at: https://github.com/ZhengPeng7/GLCNet.

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

Anomaly detection (AD), aiming to find samples that deviate from the training distribution, is essential in safety-critical applications. Though recent self-supervised learning based attempts achieve promising results by creating virtual outliers, their training objectives are less faithful to AD which requires a concentrated inlier distribution as well as a dispersive outlier distribution. In this paper, we propose Unilaterally Aggregated Contrastive Learning with Hierarchical Augmentation (UniCon-HA), taking into account both the requirements above. Specifically, we explicitly encourage the concentration of inliers and the dispersion of virtual outliers via supervised and unsupervised contrastive losses, respectively. Considering that standard contrastive data augmentation for generating positive views may induce outliers, we additionally introduce a soft mechanism to re-weight each augmented inlier according to its deviation from the inlier distribution, to ensure a purified concentration. Moreover, to prompt a higher concentration, inspired by curriculum learning, we adopt an easy-to-hard hierarchical augmentation strategy and perform contrastive aggregation at different depths of the network based on the strengths of data augmentation. Our method is evaluated under three AD settings including unlabeled one-class, unlabeled multi-class, and labeled multi-class, demonstrating its consistent superiority over other competitors.

Temporal action detection (TAD) aims to localize the start and end frames of actions in untrimmed videos, which is a challenging task due to the similarity of adjacent frames and the ambiguity of action boundaries. Previous methods often generate coarse proposals first and then perform proposal-based refinement, which is coupled with prior action detectors and leads to proposal-oriented offsets. However, this paradigm increases the training difficulty of the TAD model and is heavily influenced by the quantity and quality of the proposals. To address the above issues, we decouple the refinement process from conventional TAD methods and propose a learnable, proposal-free refinement method for fine boundary localization, named RefineTAD. We first propose a multi-level refinement module to generate multi-scale boundary offsets, score offsets and boundary-aware probability at each time point based on the feature pyramid. Then, we propose an offset focusing strategy to progressively refine the predicted results of TAD models in a coarse-to-fine manner with our multi-scale offsets. We perform extensive experiments on three challenging datasets and demonstrate that our RefineTAD significantly improves the state-of-the-art TAD methods with minimal computational overhead.

Video frame interpolation has been actively studied with the development of convolutional neural networks. However, due to the intrinsic limitations of kernel weight sharing in convolution, the interpolated frame generated by it may lose details. In contrast, the attention mechanism in Transformer can better distinguish the contribution of each pixel, and it can also capture long-range pixel dependencies, which provides great potential for video interpolation. Nevertheless, the original Transformer is commonly used for 2D images; how to develop a Transformer-based framework with consideration of temporal self-attention for video frame interpolation remains an open issue. In this paper, we propose Video Frame Interpolation Flow Transformer to incorporate motion dynamics from optical flows into the self-attention mechanism. Specifically, we design a Flow Transformer Block that calculates the temporal self-attention in a matched local area with the guidance of flow, making our framework suitable for interpolating frames with large motion while maintaining reasonably low complexity. In addition, we construct a multi-scale architecture to account for multi-scale motion, further improving the overall performance. Extensive experiments on three benchmarks demonstrate that the proposed method can generate interpolated frames with better visual quality than state-of-the-art methods.

Scene text recognition (STR) has been actively studied in recent years, with a wide range of applications in autonomous driving, image retrieval and much more. However, when a pre-trained deep STR model learns a new task, its performance on previous tasks may drop dramatically, due to catastrophic forgetting in deep neural networks. A potential solution to combat the forgetting of prior knowledge is incremental learning (IL), which has shown its effectiveness and significant progress in image classification. Yet, exploiting IL in the context of STR has been barely visited, probably because the forgetting problem is even worse in STR. To address this issue, we propose the lifelong scene text recognizer (LSTR) that learns STR tasks incrementally while alleviating forgetting. Specifically, LSTR assigns each task a set of task-specific expert modules at different stages of an STR model, while other parameters are shared among tasks. These shared parameters are only learned in the first task and remain unchanged during subsequent learning to ensure that no learned knowledge is overlooked. Moreover, in real applications, there is no prior knowledge about which task an input image belongs to, making it impossible to precisely select the corresponding expert modules. To this end, we propose the incremental task prediction network (ITPN) to identify the most related task category by pulling the features of the same task closer and pushing those of different tasks farther apart. To validate the proposed method in our newly-introduced IL setting, we collected a large-scale dataset consisting of both real and synthetic multilingual STR data. Extensive experiments on this dataset clearly show the superiority of our LSTR over state-of-the-art IL methods.

Normalization techniques are essential for accelerating the training and improving the generalization of deep neural networks (DNNs), and have successfully been used in various applications. This paper reviews and comments on the past, present and future of normalization methods in the context of DNN training. We provide a unified picture of the main motivation behind different approaches from the perspective of optimization, and present a taxonomy for understanding the similarities and differences between them. Specifically, we decompose the pipeline of the most representative normalizing activation methods into three components: the normalization area partitioning, normalization operation and normalization representation recovery. In doing so, we provide insight for designing new normalization technique. Finally, we discuss the current progress in understanding normalization methods, and provide a comprehensive review of the applications of normalization for particular tasks, in which it can effectively solve the key issues.

Learning to re-identify or retrieve a group of people across non-overlapped camera systems has important applications in video surveillance. However, most existing methods focus on (single) person re-identification (re-id), ignoring the fact that people often walk in groups in real scenarios. In this work, we take a step further and consider employing context information for identifying groups of people, i.e., group re-id. On the one hand, group re-id is more challenging than single person re-id, since it requires both a robust modeling of local individual person appearance (with different illumination conditions, pose/viewpoint variations, and occlusions), as well as full awareness of global group structures (with group layout and group member variations). On the other hand, we believe that person re-id can be greatly enhanced by incorporating additional visual context from neighboring group members, a task which we formulate as group-aware (single) person re-id. In this paper, we propose a novel unified framework based on graph neural networks to simultaneously address the above two group-based re-id tasks, i.e., group re-id and group-aware person re-id. Specifically, we construct a context graph with group members as its nodes to exploit dependencies among different people. A multi-level attention mechanism is developed to formulate both intra-group and inter-group context, with an additional self-attention module for robust graph-level representations by attentively aggregating node-level features. The proposed model can be directly generalized to tackle group-aware person re-id using node-level representations. Meanwhile, to facilitate the deployment of deep learning models on these tasks, we build a new group re-id dataset which contains more than 3.8K images with 1.5K annotated groups, an order of magnitude larger than existing group re-id datasets. Extensive experiments on the novel dataset as well as three existing datasets clearly demonstrate the effectiveness of the proposed framework for both group-based re-id tasks.

In this paper, we present a novel end-to-end group collaborative learning network, termed GCoNet+, which can effectively and efficiently (250 fps) identify co-salient objects in natural scenes. The proposed GCoNet+ achieves the new state-of-the-art performance for co-salient object detection (CoSOD) through mining consensus representations based on the following two essential criteria: 1) intra-group compactness to better formulate the consistency among co-salient objects by capturing their inherent shared attributes using our novel group affinity module (GAM); 2) inter-group separability to effectively suppress the influence of noisy objects on the output by introducing our new group collaborating module (GCM) conditioning on the inconsistent consensus. To further improve the accuracy, we design a series of simple yet effective components as follows: i) a recurrent auxiliary classification module (RACM) promoting model learning at the semantic level; ii) a confidence enhancement module (CEM) assisting the model in improving the quality of the final predictions; and iii) a group-based symmetric triplet (GST) loss guiding the model to learn more discriminative features. Extensive experiments on three challenging benchmarks, i.e., CoCA, CoSOD3k, and CoSal2015, demonstrate that our GCoNet+ outperforms the existing 12 cutting-edge models. Code has been released at https://github.com/ZhengPeng7/GCoNet_plus.

Detecting the tooth-gingiva trim line from a dental surface plays a critical role in dental treatment planning and aligner 3D printing. Existing methods treat this task as a segmentation problem, which is resolved with geometric deep learning based mesh segmentation techniques. However, these methods can only provide indirect results (i.e., segmented teeth) and suffer from unsatisfactory accuracy due to the incapability of making full use of high-resolution dental surfaces. To this end, we propose a two-stage geometric deep learning framework for automatically detecting tooth-gingiva trim lines from dental surfaces. Our framework consists of a trim line proposal network (TLP-Net) for predicting an initial trim line from the low-resolution dental surface as well as a trim line refinement network (TLR-Net) for refining the initial trim line with the information from the high-resolution dental surface. Specifically, our TLP-Net predicts the initial trim line by fusing the multi-scale features from a U-Net with a proposed residual multi-scale attention fusion module. Moreover, we propose feature bridge modules and a trim line loss to further improve the accuracy. The resulting trim line is then fed to our TLR-Net, which is a deep-based LDDMM model with the high-resolution dental surface as input. In addition, dense connections are incorporated into TLR-Net for improved performance. Our framework provides an automatic solution to trim line detection by making full use of raw high-resolution dental surfaces. Extensive experiments on a clinical dental surface dataset demonstrate that our TLP-Net and TLR-Net are superior trim line detection methods and outperform cutting-edge methods in both qualitative and quantitative evaluations.

Zero-shot learning (ZSL) aims to recognize unknown categories that are unavailable during training. Recently, generative models have shown the potential to address this challenging problem by synthesizing unseen features conditioned on semantic embeddings such as attributes. However, unidirectional generative models cannot guarantee the effective coupling between visual and semantic spaces. To this end, we propose a visual-semantic aligned bidirectional network with cycle consistency to alleviate the gap between these two spaces, generating unseen features of high quality. More importantly, we incorporate two carefully designed strategies into our bidirectional framework to improve the overall ZSL performance. Specifically, we enhance the intra-domain class divergence in both visual and semantic spaces, and in the meantime, mitigate the inter-domain shift to preserve seen-unseen domain discrimination. Experimental results on four standard benchmarks show the superiority of our framework over existing state-of-the-art methods under both conventional and generalized ZSL settings.

Image-level weakly supervised semantic segmentation (WSSS) is a fundamental yet challenging computer vision task facilitating scene understanding and automatic driving. Most existing methods resort to classification-based Class Activation Maps (CAMs) to play as the initial pseudo labels, which tend to focus on the discriminative image regions and lack customized characteristics for the segmentation task. To alleviate this issue, we propose a novel activation modulation and recalibration (AMR) scheme, which leverages a spotlight branch and a compensation branch to obtain weighted CAMs that can provide recalibration supervision and task-specific concepts. Specifically, an attention modulation module (AMM) is employed to rearrange the distribution of feature importance from the channel-spatial sequential perspective, which helps to explicitly model channel-wise interdependencies and spatial encodings to adaptively modulate segmentation-oriented activation responses. Furthermore, we introduce a cross pseudo supervision for dual branches, which can be regarded as a semantic similar regularization to mutually refine two branches. Extensive experiments show that AMR establishes a new state-of-the-art performance on the PASCAL VOC 2012 dataset, surpassing not only current methods trained with the image-level of supervision but also some methods relying on stronger supervision, such as saliency label. Experiments also reveal that our scheme is plug-and-play and can be incorporated with other approaches to boost their performance. Our code is available at: https://github.com/jieqin-ai/AMR.

Person search has recently emerged as a challenging task that jointly addresses pedestrian detection and person re-identification. Existing approaches follow a fully supervised setting where both bounding box and identity annotations are available. However, annotating identities is labor-intensive, limiting the practicability and scalability of current frameworks. This paper inventively considers weakly supervised person search with only bounding box annotations. We propose to address this novel task by investigating three levels of context clues (i.e., detection, memory and scene) in unconstrained natural images. The first two are employed to promote local and global discriminative capabilities, while the latter enhances clustering accuracy. Despite its simple design, our CGPS boosts the baseline model by 8.8% in mAP on CUHK-SYSU. Surprisingly, it even achieves comparable performance with several supervised person search models. Our code is available at https://github. com/ljpadam/CGPS.

Weakly-supervised temporal action localization (WTAL) in untrimmed videos has emerged as a practical but challenging task since only video-level labels are available. Existing approaches typically leverage off-the-shelf segment-level features, which suffer from spatial incompleteness and temporal incoherence, thus limiting their performance. In this paper, we tackle this problem from a new perspective by enhancing segment-level representations with a simple yet effective graph convolutional network, namely action complement graph network (ACGNet). It facilitates the current video segment to perceive spatial-temporal dependencies from others that potentially convey complementary clues, implicitly mitigating the negative effects caused by the two issues above. By this means, the segment-level features are more discriminative and robust to spatial-temporal variations, contributing to higher localization accuracies. More importantly, the proposed ACGNet works as a universal module that can be flexibly plugged into different WTAL frameworks, while maintaining the end-to-end training fashion. Extensive experiments are conducted on the THUMOS'14 and ActivityNet1.2 benchmarks, where the state-of-the-art results clearly demonstrate the superiority of the proposed approach.

Action segmentation, inferring temporal positions of human actions in an untrimmed video, is an important prerequisite for various video understanding tasks. Recently, unsupervised action segmentation (UAS) has emerged as a more challenging task due to the unavailability of frame-level annotations. Existing clustering- or prediction-based UAS approaches suffer from either over-segmentation or overfitting, leading to unsatisfactory results. To address those problems,we propose Predictive And Contrastive Embedding (PACE), a unified UAS framework leveraging both predictability and similarity information for more accurate action segmentation. On the basis of an auto-regressive transformer encoder, predictive embeddings are learned by exploiting the predictability of video context, while contrastive embeddings are generated by leveraging the similarity of adjacent short video clips. Extensive experiments on three challenging benchmarks demonstrate the superiority of our method, with up to 26.9% improvements in F1-score over the state of the art.

Binary optimization problems (BOPs) arise naturally in many fields, such as information retrieval, computer vision, and machine learning. Most existing binary optimization methods either use continuous relaxation which can cause large quantization errors, or incorporate a highly specific algorithm that can only be used for particular loss functions. To overcome these difficulties, we propose a novel generalized optimization method, named Alternating Binary Matrix Optimization (ABMO), for solving BOPs. ABMO can handle BOPs with/without orthogonality or linear constraints for a large class of loss functions. ABMO involves rewriting the binary, orthogonality and linear constraints for BOPs as an intersection of two closed sets, then iteratively dividing the original problems into several small optimization problems that can be solved as closed forms. To provide a strict theoretical convergence analysis, we add a sufficiently small perturbation and translate the original problem to an approximated problem whose feasible set is continuous. We not only provide rigorous mathematical proof for the convergence to a stationary and feasible point, but also derive the convergence rate of the proposed algorithm. The promising results obtained from four binary optimization tasks validate the superiority and the generality of ABMO compared with the state-of-the-art methods.

The goal of few-shot learning is to learn a classifier that can recognize unseen classes from limited support data with labels. A common practice for this task is to train a model on the base set first and then transfer to novel classes through fine-tuning or meta-learning. However, as the base classes have no overlap to the novel set, simply transferring whole knowledge from base data is not an optimal solution since some knowledge in the base model may be biased or even harmful to the novel class. In this paper, we propose to transfer partial knowledge by freezing or fine-tuning particular layer(s) in the base model. Specifically, layers will be imposed different learning rates if they are chosen to be fine-tuned, to control the extent of preserved transferability. To determine which layers to be recast and what values of learning rates for them, we introduce an evolutionary search based method that is efficient to simultaneously locate the target layers and determine their individual learning rates. We conduct extensive experiments on CUB and mini-ImageNet to demonstrate the effectiveness of our proposed method. It achieves the state-of-the-art performance on both meta-learning and non-meta based frameworks. Furthermore, we extend our method to the conventional pre-training + fine-tuning paradigm and obtain consistent improvement.

Previous studies dominantly target at self-supervised learning on real-valued networks and have achieved many promising results. However, on the more challenging binary neural networks (BNNs), this task has not yet been fully explored in the community. In this paper, we focus on this more difficult scenario: learning networks where both weights and activations are binary, meanwhile, without any human annotated labels. We observe that the commonly used contrastive objective is not satisfying on BNNs for competitive accuracy, since the backbone network contains relatively limited capacity and representation ability. Hence instead of directly applying existing self-supervised methods, which cause a severe decline in performance, we present a novel guided learning paradigm from real-valued to distill binary networks on the final prediction distribution, to minimize the loss and obtain desirable accuracy. Our proposed method can boost the simple contrastive learning baseline by an absolute gain of 5.5∼15% on BNNs. We further reveal that it is difficult for BNNs to recover the similar predictive distributions as real-valued models when training without labels. Thus, how to calibrate them is key to address the degradation in performance. Extensive experiments are conducted on the large-scale ImageNet and downstream datasets. Our method achieves substantial improvement over the simple contrastive learning baseline, and is even comparable to many mainstream supervised BNN methods. Code is available at https://github.com/szq0214/S2-BNN.

Person search aims to simultaneously localize and identify a query person from realistic, uncropped images, which can be regarded as the unified task of pedestrian detection and person re-identification (re-id). Most existing works employ two-stage detectors like Faster-RCNN, yielding encouraging accuracy but with high computational overhead. In this work, we present the Feature-Aligned Person Search Network (AlignPS), the first anchor-free framework to efficiently tackle this challenging task. AlignPS explicitly addresses the major challenges, which we summarize as the misalignment issues in different levels (i.e., scale, region, and task), when accommodating an anchor-free detector for this task. More specifically, we propose an aligned feature aggregation module to generate more discriminative and robust feature embeddings by following a "re-id first" principle. Such a simple design directly improves the baseline anchor-free model on CUHK-SYSU by more than 20% in mAP. Moreover, AlignPS outperforms state-of-the-art two-stage methods, with a higher speed. The code is available at https://github.com/daodaofr/AlignPS.

In this work, we propose a two-stage framework to efficiently and effectively detect actions on-the-fly. An action location network (ALN) is developed in the first stage to judge whether the current frame is action-related, while the second stage involves an action classification network (ACN) to further identify the action category. In this way, irrelevant negative frames are quickly discarded and actions are detected as early as they occur. Moreover, we highlight human areas at both the stages by respectively incorporating a human detector and a human mask layer. As a result, more accurate spatial-temporal windows of actions are detected, based on which more robust features are extracted for classification. Experimental results on two popular benchmarks demonstrate the superior performance of the proposed approach.

Accurately describing and detecting 2D and 3D key-points is crucial to establishing correspondences across images and point clouds. Despite a plethora of learning-based 2D or 3D local feature descriptors and detectors having been proposed, the derivation of a shared descriptor and joint keypoint detector that directly matches pixels and points remains under-explored by the community. This work takes the initiative to establish fine-grained correspondences between 2D images and 3D point clouds. In order to directly match pixels and points, a dual fully-convolutional framework is presented that maps 2D and 3D inputs into a shared latent representation space to simultaneously describe and detect keypoints. Furthermore, an ultra-wide reception mechanism and a novel loss function are designed to mitigate the intrinsic information variations between pixel and point local regions. Extensive experimental results demonstrate that our framework shows competitive performance in fine-grained matching between images and point clouds and achieves state-of-the-art results for the task of indoor visual localization. Our source code is available at https://github.com/BingCS/P2-Net.

Action detection plays an important role in high-level video understanding and media interpretation. Many existing studies fulfill this spatio-temporal localization by modeling the context, capturing the relationship of actors, objects, and scenes conveyed in the video. However, they often universally treat all the actors without considering the consistency and distinctness between individuals, leaving much room for improvement. In this paper, we explicitly highlight the identity information of the actors in terms of both long-term and short-term context through a graph memory network, namely identity-aware graph memory network (IGMN). Specifically, we propose the hierarchical graph neural network (HGNN) to comprehensively conduct long-term relation modeling within the same identity as well as between different ones. Regarding short-term context, we develop a dual attention module (DAM) to generate identity-aware constraint to reduce the influence of interference by the actors of different identities. Extensive experiments on the challenging AVA dataset demonstrate the effectiveness of our method, which achieves state-of-the-art results on AVA v2.1 and v2.2.

Visual storytelling aims to automatically generate a human-like short story given an image stream. Most existing works utilize either scene-level or object-level representations, neglecting the interaction among objects in each image and the sequential dependency between consecutive images. In this paper, we present a novel Latent Memory-augmented Graph Transformer~(LMGT ), a Transformer based framework for visual story generation. LMGT directly inherits the merits from the Transformer, which is further enhanced with two carefully designed components, i.e., a graph encoding module and a latent memory unit. Specifically, the graph encoding module exploits the semantic relationships among image regions and attentively aggregates critical visual features based on the parsed scene graphs. Furthermore, to better preserve inter-sentence coherence and topic consistency, we introduce an augmented latent memory unit that learns and records highly summarized latent information as the story line from the image stream and the sentence history. Experimental results on three widely-used datasets demonstrate the superior performance of LMGT over the state-of-the-art methods.

Recent advances in self-supervised learning with instance-level contrastive objectives facilitate unsupervised clustering. However, a standalone datum is not perceiving the context of the holistic cluster, and may undergo sub-optimal assignment. In this paper, we extend the mainstream contrastive learning paradigm to a cluster-level scheme, where all the data subjected to the same cluster contribute to a unified representation that encodes the context of each data group. Contrastive learning with this representation then rewards the assignment of each datum. To implement this vision, we propose twin-contrast clustering (TCC). We define a set of categorical variables as clustering assignment confidence, which links the instance-level learning track with the cluster-level one. On one hand, with the corresponding assignment variables being the weight, a weighted aggregation along the data points implements the set representation of a cluster. We further propose heuristic cluster augmentation equivalents to enable cluster-level contrastive learning. On the other hand, we derive the evidence lower-bound of the instance-level contrastive objective with the assignments. By reparametrizing the assignment variables, TCC is trained end-to-end, requiring no alternating steps. Extensive experiments show that TCC outperforms the state-of-the-art on benchmarked datasets.

With the current exponential growth of video-based social networks, video retrieval using natural language is receiving ever-increasing attention. Most existing approaches tackle this task by extracting individual frame-level spatial features to represent the whole video, while ignoring visual pattern consistencies and intrinsic temporal relationships across different frames. Furthermore, the semantic correspondence between natural language queries and person-centric actions in videos has not been fully explored. To address these problems, we propose a novel binary representation learning framework, named Semantics-aware Spatial-temporal Binaries ( [Formula: see text]Bin), which simultaneously considers spatial-temporal context and semantic relationships for cross-modal video retrieval. By exploiting the semantic relationships between two modalities, [Formula: see text]Bin can efficiently and effectively generate binary codes for both videos and texts. In addition, we adopt an iterative optimization scheme to learn deep encoding functions with attribute-guided stochastic training. We evaluate our model on three video datasets and the experimental results demonstrate that [Formula: see text]Bin outperforms the state-of-the-art methods in terms of various cross-modal video retrieval tasks.

Trajectory forecasting and imputation are pivotal steps towards understanding the movement of human and objects, which are quite challenging since the future trajectories and missing values in a temporal sequence are full of uncertainties, and the spatial-temporally contextual correlation is hard to model. Yet, the relevance between sequence prediction and imputation is disregarded by existing approaches. To this end, we propose a novel imitative non-autoregressive modeling method to simultaneously handle the trajectory prediction task and the missing value imputation task. Specifically, our framework adopts an imitation learning paradigm, which contains a recurrent conditional variational autoencoder (RC-VAE) as a demonstrator, and a non-autoregressive transformation model (NART) as a learner. By jointly optimizing the two models, RC-VAE can predict the future trajectory and capture the temporal relationship in the sequence to supervise the NART learner. As a result, NART learns from the demonstrator and imputes the missing value in a non autoregressive strategy. We conduct extensive experiments on three popular datasets, and the results show that our model achieves state-of-the-art performance across all the datasets.

Conventional unsupervised hashing methods usually take advantage of similarity graphs, which are either pre-computed in the high-dimensional space or obtained from random anchor points. On the one hand, existing methods uncouple the procedures of hash function learning and graph construction. On the other hand, graphs empirically built upon original data could introduce biased prior knowledge of data relevance, leading to sub-optimal retrieval performance. In this paper, we tackle the above problems by proposing an efficient and adaptive code-driven graph, which is updated by decoding in the context of an auto-encoder. Specifically, we introduce into our framework twin bottlenecks (i.e., latent variables) that exchange crucial information collaboratively. One bottleneck (i.e., binary codes) conveys the high-level intrinsic data structure captured by the code-driven graph to the other (i.e., continuous variables for low-level detail information), which in turn propagates the updated network feedback for the encoder to learn more discriminative binary codes. The auto-encoding learning objective literally rewards the code-driven graph to learn an optimal encoder. Moreover, the proposed model can be simply optimized by gradient descent without violating the binary constraints. Experiments on benchmarked datasets clearly show the superiority of our framework over the state-of-the-art hashing methods. Our source code can be found at https://github.com/ymcidence/TBH.

Video-based person re-identification (re-ID) is an important research topic in computer vision. The key to tackling the challenging task is to exploit both spatial and temporal clues in video sequences. In this work, we propose a novel graph-based framework, namely Multi-Granular Hypergraph (MGH), to pursue better representational capabilities by modeling spatiotemporal dependencies in terms of multiple granularities. Specifically, hypergraphs with different spatial granularities are constructed using various levels of part-based features across the video sequence. In each hypergraph, different temporal granularities are captured by hyperedges that connect a set of graph nodes (i.e., part-based features) across different temporal ranges. Two critical issues (misalignment and occlusion) are explicitly addressed by the proposed hypergraph propagation and feature aggregation schemes. Finally, we further enhance the overall video representation by learning more diversified graph-level representations of multiple granularities based on mutual information minimization. Extensive experiments on three widely-adopted benchmarks clearly demonstrate the effectiveness of the proposed framework. Notably, 90.0% top-1 accuracy on MARS is achieved using MGH, outperforming the state-of-the-arts.

The Deep Convolutional Neural Networks (CNNs) have obtained a great success for pattern recognition, such as recognizing the texts in images. But existing CNNs based frameworks still have several drawbacks: 1) the traditaional pooling operation may lose important feature information and is unlearnable; 2) the tradi-tional convolution operation optimizes slowly and the hierar-chical features from different layers are not fully utilized. In this work, we address these problems by developing a novel deep network model called Fully-Convolutional Intensive Feature Flow Neural Network (IntensiveNet). Specifically, we design a further dense block called intensive block to extract the feature information, where the original inputs and two dense blocks are connected tightly. To encode data appropriately, we present the concepts of dense fusion block and further dense fusion opera-tions for our new intensive block. By adding short connections to different layers, the feature flow and coupling between layers are enhanced. We also replace the traditional convolution by depthwise separable convolution to make the operation efficient. To prevent important feature information being lost to a certain extent, we use a convolution operation with stride 2 to replace the original pooling operation in the customary transition layers. The recognition results on large-scale Chinese string and MNIST datasets show that our IntensiveNet can deliver enhanced recog-nition results, compared with other related deep models.

Person re-identification (ReID) has recently received extensive research interests due to its diverse applications in multimedia analysis and computer vision. However, the majority of existing works focus on improving matching accuracy, while ignoring matching efficiency. In this work, we present a novel binary representation learning framework for efficient person ReID, namely Deep Local Binary Coding (DLBC). Different from existing deep binary ReID approaches, DLBC attempts to learn discriminative binary codes by explicitly interacting with local visual details. Specifically, DLBC first extracts a set of local features from spatially salient regions of pedestrian images. Subsequently, DLBC formulates a new binary-local semantic mutual information (BSMI) maximization term, based on which a self-lifting (SL) block is built to further exploit the semantic importance of local features. The BSMI term together with the SL block simultaneously enhances the dependency of binary codes on selected local features as well as their robustness to cross-view visual inconsistency. In addition, an efficient optimizing method is developed to train the proposed deep models with orthogonal and binary constraints. Extensive experiments reveal that DLBC significantly minimizes the accuracy gap between binary ReID methods and the state-of-the-art real-valued ones, whilst remarkably reducing query time and memory cost.

In the era of big data, few-shot learning has recently received much attention in multimedia analysis and computer vision due to its appealing ability of learning from scarce labeled data. However, it has been largely underdeveloped in the video domain, which is even more challenging due to the huge spatial-temporal variability of video data. In this paper, we address few-shot video classification by learning an ensemble of SlowFast networks augmented with memory units. Specifically, we introduce a family of few-shot learners based on SlowFast networks which are used to extract informative features at multiple rates, and we incorporate a memory unit into each network to enable encoding and retrieving crucial information instantly. Furthermore, we propose a choice controller network to leverage the diversity of few-shot learners by learning to adaptively assign a confidence score to each SlowFast memory network, leading to a strong classifier for enhanced prediction. Experimental results on two widely-adopted video datasets demonstrate the effectiveness of the proposed method, as well as its superior performance over the state-of-the-art approaches.

Sports video analysis has received increasing attention in recent years. Athlete tracking and action recognition are its two major issues that are highly related to each other; however, they are individually considered and processed in the existing studies. In this paper, we propose a joint framework for athlete tracking and action recognition in sports videos. In athlete tracking, we propose a scaling and occlusion robust tracker, named scaling and occlusion robust compressive tracking (CT), to localize the position of specific athlete in each frame. It follows the approach of CT but extends it in two aspects, i.e., scale refinement as well as occlusion recovery. For the former, an objectness method, edge box, is adopted to generate proposals, which replace the fixed sampling boxes in CT and better fit the scales of the candidate objects. For the latter, a candidate obstruction-based solution is presented, which brings in additional trackers to detect possible obstructions and to relocate the target as occlusion ends. Regarding action recognition, we propose a long-term recurrent region-guided convolutional network, which recognizes pre-defined actions by modeling discriminative temporal cues of the tracking results. We employ SPP-net to extract the robust feature of the tracked region of each frame. The features of all the frames are then fed into a stack of recurrent sequence models to capture the long-term region-level information. We extensively evaluate the proposed approach on a newly collected sports video benchmark and on the off-the-shelf UIUC2 dataset, and the experimental results clearly show its effectiveness.

In real life, group activity recognition plays a significant and fundamental role in a variety of applications, e.g. sports video analysis, abnormal behavior detection, and intelligent surveillance. In a complex dynamic scene, a crucial yet challenging issue is how to better model the spatio-temporal contextual information and inter-person relationship. In this paper, we present a novel attentive semantic recurrent neural network (RNN), namely, stagNet, for understanding group activities and individual actions in videos, by combining the spatio-temporal attention mechanism and semantic graph modeling. Specifically, a structured semantic graph is explicitly modeled to express the spatial contextual content of the whole scene, which is further incorporated with the temporal factor through structural-RNN. By virtue of the “factor sharing” and “message passing” mechanisms, our stagNet is capable of extracting discriminative and informative spatio-temporal representations and capturing inter-person relationships. Moreover, we adopt a spatio-temporal attention model to focus on key persons/frames for improved recognition performance. Besides, a body-region attention and a global-part feature pooling strategy are devised for individual action recognition. In experiments, four widely-used public datasets are adopted for performance evaluation, and the extensive results demonstrate the superiority and effectiveness of our method.

Zero-shot learning (ZSL) is a challenging task due to the lack of unseen class data during training. Existing works attempt to establish a mapping between the visual and class spaces through a common intermediate semantic space. The main limitation of existing methods is the strong bias towards seen class, known as the domain shift problem, which leads to unsatisfactory performance in both conventional and generalized ZSL tasks. To tackle this challenge, we propose to convert ZSL to the conventional supervised learning by generating features for unseen classes. To this end, a joint generative model that couples variational autoencoder (VAE) and generative adversarial network (GAN), called Zero-VAE-GAN, is proposed to generate high-quality unseen features. To enhance the class-level discriminability, an adversarial categorization network is incorporated into the joint framework. Besides, we propose two self-training strategies to augment unlabeled unseen features for the transductive extension of our model, addressing the domain shift problem to a large extent. Experimental results on five standard benchmarks and a large-scale dataset demonstrate the superiority of our generative model over the state-of-the-art methods for conventional, especially generalized ZSL tasks. Moreover, the further improvement of the transductive setting demonstrates the effectiveness of the proposed self-training strategies.

Both interclass variances and intraclass similarities are crucial for improving the classification performance of discriminative dictionary learning (DDL) algorithms. However, existing DDL methods often ignore the combination between the interclass and intraclass properties of dictionary atoms and coding coefficients. To address this problem, in this paper, we propose a discriminative Fisher embedding dictionary learning (DFEDL) algorithm that simultaneously establishes Fisher embedding models on learned atoms and coefficients. Specifically, we first construct a discriminative Fisher atom embedding model by exploring the Fisher criterion of the atoms, which encourages the atoms of the same class to reconstruct the corresponding training samples as much as possible. At the same time, a discriminative Fisher coefficient embedding model is formulated by imposing the Fisher criterion on the profiles (row vectors of the coding coefficient matrix) and coding coefficients, which forces the coding coefficient matrix to become a block-diagonal matrix. Since the profiles can indicate which training samples are represented by the corresponding atoms, the proposed two discriminative Fisher embedding models can alternatively and interactively promote the discriminative capabilities of the learned dictionary and coding coefficients. The extensive experimental results demonstrate that the proposed DFEDL algorithm achieves superior performance in comparison with some state-of-the-art dictionary learning algorithms on both hand-crafted and deep learning-based features.

Sketch-based 3D shape retrieval has been extensively studied in recent works, most of which focus on improving the retrieval accuracy, whilst neglecting the efficiency. In this paper, we propose a novel framework for efficient sketch-based 3D shape retrieval, i.e., Deep Sketch-Shape Hashing (DSSH), which tackles the challenging problem from two perspectives. Firstly, we propose an intuitive 3D shape representation method to deal with unaligned shapes with arbitrary poses. Specifically, the proposed Segmented Stochastic-viewing Shape Network models discriminative 3D representations by a set of 2D images rendered from multiple views, which are stochastically selected from non-overlapping spatial segments of a 3D sphere. Secondly, Batch-Hard Binary Coding (BHBC) is developed to learn semantics-preserving compact binary codes by mining the hardest samples. The overall framework is jointly learned by developing an alternating iteration algorithm. Extensive experimental results on three benchmarks show that DSSH improves both the retrieval efficiency and accuracy remarkably, compared to the state-of-the-art methods.

In recent years, scene parsing has captured increasing attention in computer vision. Previous works have demonstrated promising performance in this task. However, they mainly utilize holistic features, whilst neglecting the rich semantic knowledge and inter-object relationships in the scene. In addition, these methods usually require a large number of pixel-level annotations, which is too expensive in practice. In this paper, we propose a novel Knowledge Embedded Generative Adversarial Networks, dubbed as KE-GAN, to tackle the challenging problem in a semi-supervised fashion. KE-GAN captures semantic consistencies of different categories by devising a Knowledge Graph from the large-scale text corpus. In addition to readily-available unlabeled data, we generate synthetic images to unveil rich structural information underlying the images. Moreover, a pyramid architecture is incorporated into the discriminator to acquire multi-scale contextual information for better parsing results. Extensive experimental results on four standard benchmarks demonstrate that KE-GAN is capable of improving semantic consistencies and learning better representations for scene parsing, resulting in the state-of-the-art performance.

Zero-shot learning (ZSL) aims to classify images from unseen categories, by merely utilizing seen class images as the training data. Existing works on ZSL mainly leverage the global features or learn the global regions, from which, to construct the embeddings to the semantic space. However, few of them study the discrimination power implied in local image regions (parts), which, in some sense, correspond to semantic attributes, have stronger discrimination than attributes, and can thus assist the semantic transfer between seen/unseen classes. In this paper, to discover (semantic) regions, we propose the attentive region embedding network (AREN), which is tailored to advance the ZSL task. Specifically, AREN is end-to-end trainable and consists of two network branches, i.e., the attentive region embedding (ARE) stream, and the attentive compressed second-order embedding (ACSE) stream. ARE is capable of discovering multiple part regions under the guidance of the attention and the compatibility loss. Moreover, a novel adaptive thresholding mechanism is proposed for suppressing redundant (such as background) attention regions. To further guarantee more stable semantic transfer from the perspective of second-order collaboration, ACSE is incorporated into the AREN. In the comprehensive evaluations on four benchmarks, our models achieve state-of-the-art performances under ZSL setting, and compelling results under generalized ZSL setting.

Person re-identification (ReID) aims at associating persons with the same identity across different views/scenes. Most existing methods improve matching accuracy by proposing high-dimensional real-valued features to represent person images comprehensively. However, considering the increasing data scale in real-world applications, the storage and matching efficiencies should be paid attention to as well. In this paper, we propose a binary coding approach for efficient ReID, inspired by the recent advances in adversarial learning. Specifically, the proposed Adversarial Binary Coding (ABC) implicitly fits the feature distribution to the expected binary one by optimizing the Wasserstein distance. To further enhance the semantic discriminability of binary codes, we seamlessly embed the ABC into a similarity measuring deep neural network. By end-to-end learning the framework, compact and discriminative binary features are generated for efficient and accurate ReID. Extensive experiments on large-scale benchmarks demonstrate the superiority of our approach over the state-of-the-art methods in both efficiency and accuracy.

We propose a novel structured discriminative block- diagonal dictionary learning method, referred to as scalable Locality-Constrained Projective Dictionary Learning (LC-PDL), for efficient representation and classification. To improve the scalability by saving both training and testing time, our LC-PDL aims at learning a structured discriminative dictionary and a block-diagonal representation without using costly l0/l1-norm. Besides, it avoids extra time-consuming sparse reconstruction process with the well-trained dictionary for new sample as many existing models. More importantly, LC-PDL avoids using the com- plementary data matrix to learn the sub-dictionary over each class. To enhance the performance, we incorporate a locality constraint of atoms into the DL procedures to keep local information and obtain the codes of samples over each class separately. A block-diagonal discriminative approximation term is also derived to learn a discriminative projection to bridge data with their codes by extracting the special block-diagonal features from data, which can ensure the approximate coefficients to associate with its label information clearly. Then, a robust multiclass classifier is trained over extracted block-diagonal codes for accurate label predictions. Experimental results verify the effectiveness of our algorithm.

In this paper, we propose a novel unsupervised Nonnegative Adaptive Feature Extraction (NAFE) algorithm for data representation and classification. The formulation of NAFE integrates the sparsity constrained nonnegative matrix factorization (NMF), representation learning, and adaptive reconstruction weight learning into a unified model. Specifically, NAFE performs feature and weight learning over the new robust representations of NMF for more accurate measure and representation. For nonnegative adaptive feature extraction, our NAFE first utilizes the sparsity constrained NMF to obtain the new and robust representations of the original data. To preserve the manifold structures of the learnt new representations, we also incorporate a neighborhood reconstruction error over the weight matrix for joint minimization. Note that to further improve the representation power, the weights are jointly shared in the new low-dimensional nonnegative representation space, low-dimensional nonlinear manifold space, and low-dimensional projective subspace, i.e., local neighborhood information is clearly preserved in different feature spaces so that informative representations and features can be jointly obtained. To enable NAFE to extract features from new data, we also include a feature approximation error by a linear projection so that the learnt extractor can obtain features from new data efficiently. Extensive simulations show that our formulation can deliver state-of-the-art results on several public databases for feature extraction and classification, compared with several related methods.

Group activity recognition in videos is a challenging task, with two major issues, i.e. attending to those persons and their body parts that contribute significantly to the activity, and modeling contextual person structures in the group. Most previous approaches fail to provide a practical solution to jointly address both issues, however. In this paper, we propose to simultaneously deal with both issues via a hierarchical attention and context modeling framework based on Long Short-Term Memory (LSTM) networks. For the former, we propose `Hierarchical Attention Networks' applied at the part/person level, capable of attending distinctively to different persons and their body parts. For the latter, we build `Hierarchical Context Networks' that take the attentively pooled person-level features as input and recurrently model intra/inter-group contextual structures. The attentive and contextual representations are concatenated and fed into another LSTM to generate high-level discriminative temporal representations for group activity recognition. Extensive experiments on two widely-used group activity datasets demonstrate the effectiveness and superiority of the proposed framework.

Most existing low-rank sparse embedding models extract features of data in the original input space and usually separate the manifold preservation step from the coding process, which may result in the decreased performance. In this paper, a novel Robust Adaptive Low-rank and Sparse Embedding (RALSE) framework is technically proposed for salient feature extraction of the high-dimensional data by seamlessly integrating the joint low-rank and sparse recovery with the robust adaptive salient feature extraction. Specifically, our RALSE integrates the joint low-rank and sparse representation, adaptive neighborhood preserving graph weight learning and the robustness-promoting representation into a unified framework. For accurate similarity measure, RALSE computes the adaptive weights by minimizing the reconstruction error over the noise-removed data and salient features simultaneously, where L1-norm is regularized to ensure the sparse properties of learnt weights. RALSE can also ensure the learnt projection to preserve local neighborhood information of embedded features clearly and adaptively. The projection is not only modeled under joint low-rank and sparse regularization, but also computed from a clean subspace, making it powerful for the salient feature extraction. Thus, the learnt low-rank sparse features would be more accurate for subsequent classification. Extensive results demonstrate the effectiveness of our RALSE formulation for data representation and classification.

We mainly propose a robust Embedded Locality-Constrained Label Consistent Dictionary Learning (ELC2DL) framework for discriminative classification. ELC2DL improves the representation and classification performance by performing DL in the noise-removed sparse embedding space, since most real data often contains noise and performing DL over noisy data for reconstruction may decrease performance potentially. To reduce the noise in data, our model computes a sparse projection jointly for noise reduction and then uses the noise-removed data for DL. By incorporating a noise-reduction term with a discriminative locality-constrained label consistent term that associates the label information with each dictionary atom to preserve local structure of training data, a noise-reduction projection, an over-complete dictionary and discriminative sparse codes are obtained jointly. Simulations on several image databases show that our algorithm can deliver enhanced performance over other state-of-the-arts.

In this paper, we propose an analysis mechanism-based structured analysis discriminative dictionary learning analysis discriminative dictionary learning, framework. The ADDL seamlessly integrates ADDL, analysis representation, and analysis classifier training into a unified model. The applied analysis mechanism can make sure that the learned dictionaries, representations, and linear classifiers over different classes are independent and discriminating as much as possible. The dictionary is obtained by minimizing a reconstruction error and an analytical incoherence promoting term that encourages the subdictionaries associated with different classes to be independent. To obtain the representation coefficients, ADDL imposes a sparse -norm constraint on the coding coefficients instead of using or norm, since the - or -norm constraint applied in most existing DL criteria makes the training phase time consuming. The code-extraction projection that bridges data with the sparse codes by extracting special features from the given samples is calculated via minimizing a sparse code approximation term. Then we compute a linear classifier based on the approximated sparse codes by an analysis mechanism to simultaneously consider the classification and representation powers. Thus, the classification approach of our model is very efficient, because it can avoid the extra time-consuming sparse reconstruction process with trained dictionary for each new test data as most existing DL algorithms. Simulations on real image databases demonstrate that our ADDL model can obtain superior performance over other state of the arts.

We propose a robust inductive semi-supervised label prediction model over the embedded representation, termed adaptive embedded label propagation with weight learning (AELP-WL), for classification. AELP-WL offers several properties. First, our method seamlessly integrates the robust adaptive embedded label propagation with adaptive weight learning into a unified framework. By minimizing the reconstruction errors over embedded features and embedded soft labels jointly, our AELP-WL can explicitly ensure the learned weights to be joint optimal for representation and classification, which differs from most existing LP models that perform weight learning separately by an independent step before label prediction. Second, existing models usually precalculate the weights over the original samples that may contain unfavorable features and noise decreasing performance. To this end, our model adds a constraint that decomposes original data into a sparse component encoding embedded noise-removed sparse representations of samples and a sparse error part fitting noise, and then performs the adaptive weight learning over the embedded sparse representations. Third, our AELP-WL computes the projected soft labels by trading-off the manifold smoothness and label fitness errors over the adaptive weights and the embedded representations for enhancing the label estimation power. By including a regressive label approximation error for simultaneous minimization to correlate sample features with the embedded soft labels, the out-of-sample issue is naturally solved. By minimizing the reconstruction errors over features and embedded soft labels, classification error and label approximation error jointly, state-of-the-art results are delivered.

Numerous methods have been proposed for person re-identification, most of which however neglect the matching efficiency. Recently, several hashing based approaches have been developed to make re-identification more scalable for large-scale gallery sets. Despite their efficiency, these works ignore cross-camera variations, which severely deteriorate the final matching accuracy. To address the above issues, we propose a novel hashing based method for fast person re-identification, namely Cross-camera Semantic Binary Transformation (CSBT). CSBT aims to transform original high-dimensional feature vectors into compact identity-preserving binary codes. To this end, CSBT first employs a subspace projection to mitigate cross-camera variations, by maximizing intra-person similarities and inter-person discrepancies. Subsequently, a binary coding scheme is proposed via seamlessly incorporating both the semantic pairwise relationships and local affinity information. Finally, a joint learning framework is proposed for simultaneous subspace projection learning and binary coding based on discrete alternating optimization. Experimental results on four benchmarks clearly demonstrate the superiority of CSBT over the state-of-the-art methods.

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

Recently, zero-shot action recognition (ZSAR) has emerged with the explosive growth of action categories. In this paper, we explore ZSAR from a novel perspective by adopting the Error-Correcting Output Codes (dubbed ZSECOC). Our ZSECOC equips the conventional ECOC with the additional capability of ZSAR, by addressing the domain shift problem. In particular, we learn discriminative ZSECOC for seen categories from both category-level semantics and intrinsic data structures. This procedure deals with domain shift implicitly by transferring the well-established correlations among seen categories to unseen ones. Moreover, a simple semantic transfer strategy is developed for explicitly transforming the learned embeddings of seen categories to better fit the underlying structure of unseen categories. As a consequence, our ZSECOC inherits the promising characteristics from ECOC as well as overcomes domain shift, making it more discriminative for ZSAR. We systematically evaluate ZSECOC on three realistic action benchmarks, i.e. Olympic Sports, HMDB51 and UCF101. The experimental results clearly show the superiority of ZSECOC over the state-of-the-art methods.

In this letter, we propose a novel approach for learning semantics-driven attributes, which are discriminative for zero-shot visual recognition. Latent attributes are derived in a principled manner, aiming at maintaining class-level semantic relatedness and attribute-wise balancedness. Unlike existing methods that binarize learned real-valued attributes via a quantization stage, we directly learn the optimal binary attributes by effectively addressing a discrete optimization problem. Particularly, we propose a class-wise discrete descent algorithm, based on which latent attributes of each class are learned iteratively. Moreover, we propose to simultaneously predict multiple attributes from low-level features via multioutput neural networks (MONN), which can model intrinsic correlation among attributes and make prediction more tractable. Extensive experiments on two standard datasets clearly demonstrate the superiority of our method over the state-of-the-arts.

Human action recognition in videos has been extensively studied in recent years due to its wide range of applications. Instead of classifying video sequences into a number of action categories, in this paper, we focus on a particular problem of action similarity labeling (ASLAN), which aims at verifying whether a pair of videos contain the same type of action or not. To address this challenge, a novel approach called compressive sequential learning (CSL) is proposed by leveraging the compressive sensing theory and sequential learning. We first project data points to a low-dimensional space by effectively exploring an important property in compressive sensing: the restricted isometry property. In particular, a very sparse measurement matrix is adopted to reduce the dimensionality efficiently. We then learn an ensemble classifier for measuring similarities between pairwise videos by iteratively minimizing its empirical risk with the AdaBoost strategy on the training set. Unlike conventional AdaBoost, the weak learner for each iteration is not explicitly defined and its parameters are learned through greedy optimization. Furthermore, an alternative of CSL named compressive sequential encoding is developed as an encoding technique and followed by a linear classifier to address the similarity-labeling problem. Our method has been systematically evaluated on four action data sets: ASLAN, KTH, HMDB51, and Hollywood2, and the results show the effectiveness and superiority of our method for ASLAN.

Learning based hashing methods, which aim at learning similarity-preserving binary codes for efficient nearest neighbor search, have been actively studied recently. A majority of the approaches address hashing problems for image collections. However, due to the extra temporal information, videos are usually represented by much higher dimensional (thousands or even more) features compared with images, causing high computational complexity for conventional hashing schemes. In this paper, we propose a simple and efficient hashing scheme for high-dimensional video data. This method, called Disaggregation Hashing, exploits the correlations among different feature dimensions. An intuitive feature disaggregation method is first proposed, followed by a novel hashing algorithm based on different feature clusters. We demonstrate the efficiency and effectiveness of our method by theoretical analysis and exploring its application on action retrieval from video databases. Extensive experiments show the superiority of our binary coding scheme over state-of-the-art hashing methods.

The local space-time feature is an effective way to represent video data and achieves state-of-the-art performance in action recognition. However, in majority of cases, it only captures the static or dynamic cues of the image sequence. In this paper, we propose a novel kinematic descriptor, namely Static and Dynamic fEature Velocity (SDEV), which models the changes of both static and dynamic information with time for action recognition. It is not only discriminative itself, but also complementary to the existing descriptors, thus leading to more comprehensive representation of actions by their combination. Evaluated on two public databases, i.e. UCF sports and Olympic Sports, the results clearly illustrate the competency of SDEV.

Object Classification in traffic scene surveillance has gained popularity in recent years. Traditional methods tend to utilize a large number of labeled training samples to achieve a satisfactory classification performance. However, labels of samples are not always available and manual labeling work is both time and labor consuming. To address the problem, a large number of semi-supervised learning based methods have been proposed, but most of them only focus on the offline settings. Motivated by an active learning framework, a novel online learning strategy is proposed in this paper. Furthermore, an intuitive semi-supervised learning method, which incorporates the spirits of both the online and active learning, is proposed and utilized in the scenario of traffic scene surveillance. The proposed learning framework is evaluated on the BUAA-IRIP traffic database, and the observed superior performance proves the effectiveness of our approach.

Human action recognition is a hot topic in computer vision field. Various applicable approaches have been proposed to recognize different types of actions. However, the recognition performance deteriorates rapidly when the viewpoint changes. Traditional approaches aim to address the problem by inductive transfer learning, in which target-view samples are manually labeled. In this paper, we present a novel approach for cross-view action recognition based on transductive transfer learning. We address the problem by transferring instances across views. In our settings, both labels of examples from the target view and the corresponding relation between examples from pairwise views are dispensable. Experimental results on the IXMAS multi-view data set demonstrate the effectiveness of our approach, and are comparable to the state of the art.