Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Guangming ShiSuggested from co-authorship
- Suggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works87 from public data
- Generalizing to Unseen Domains: A Survey on Domain Generalization1,906
This paper provides a formal definition of domain generalization and discusses several related fields, and categorizes recent algorithms into three classes: data manipulation, representation learning, and learning strategy, and presents several popular algorithms in detail for each category.
- An End-to-End Spatio-Temporal Attention Model for Human Action Recognition from Skeleton Data1,071
This work proposes an end-to-end spatial and temporal attention model for human action recognition from skeleton data on top of the Recurrent Neural Networks with Long Short-Term Memory (LSTM), which learns to selectively focus on discriminative joints of skeleton within each frame of the inputs and pays different levels of attention to the outputs of different frames.
- Co-Occurrence Feature Learning for Skeleton Based Action Recognition Using Regularized Deep LSTM Networks930
This work takes the skeleton as the input at each time slot and introduces a novel regularization scheme to learn the co-occurrence features of skeleton joints, and proposes a new dropout algorithm which simultaneously operates on the gates, cells, and output responses of the LSTM neurons.
- View Adaptive Recurrent Neural Networks for High Performance Human Action Recognition from Skeleton Data675
A novel view adaptation scheme to automatically regulate observation viewpoints during the occurrence of an action by design a view adaptive recurrent neural network with LSTM architecture, which enables the network itself to adapt to the most suitable observation viewpoints from end to end.
- Relation-Aware Global Attention for Person Re-Identification652
This work proposes an effective Relation-Aware Global Attention (RGA) module which captures the global structural information for better attention learning and proposes to stack the relations, i.e., its pairwise correlations/affinities with all the feature positions together to learn the attention with a shallow convolutional model.
- Semantics-Guided Neural Networks for Efficient Skeleton-Based Human Action Recognition647
A simple yet effective semantics-guided neural network (SGN) for skeleton-based action recognition that achieves the state-of-the-art performance on the NTU60, NTU120, and SYSU datasets.
- View Adaptive Neural Networks for High Performance Skeleton-Based Human Action Recognition522
A novel view adaptation scheme, which automatically determines the virtual observation viewpoints over the course of an action in a learning based data driven manner, and a two-stream scheme (referred to as VA-fusion) that fuses the scores of the two networks to provide the final prediction, obtaining enhanced performance.
- Style Normalization and Restitution for Generalizable Person Re-Identification426
The aim of this paper is to design a generalizable person ReID framework which trains a model on source domains yet is able to generalize/perform well on target domains, and to enforce a dual causal loss constraint in SNR to encourage the separation of identity-relevant features and identity-irrelevant features.
- Densely Semantically Aligned Person Re-Identification302
This work is the first to make use of fine grained semantics for addressing misalignment problems for re-ID and construct a set of densely semantically aligned part images (DSAP-images), where the same spatial positions have the same semantics across different person images.
- 257
- Online Human Action Detection Using Joint Classification-Regression Recurrent Neural Networks226
This paper proposes a multi-task end-to-end Joint Classification-Regression Recurrent Neural Network to better explore the action type and temporal localization information and demonstrates very promising performance of this scheme.
- Diffusion Models for Image Restoration and Enhancement: A Comprehensive Survey224
This first to present a comprehensive review of recent diffusion model-based methods on image restoration, encompassing the learning paradigm, conditional strategy, framework design, modeling strategy, and evaluation, and proposes nine potential and challenging directions for the future research of diffusion model-based IR.
- ReSTR: Convolution-free Referring Image Segmentation Using Transformers210
The first convolution-free model for referring image segmentation using transformers, dubbed ReSTR, is presented, which can capture long-range dependencies between entities within each modality and outperforms all existing models.
- Exploiting Sample Uncertainty for Domain Adaptive Person Re-Identification194
This paper proposes to estimate and exploit the credibility of the assigned pseudo-label of each sample to alleviate the influence of noisy labels, by suppressing the contribution of noisy samples, and introduces uncertainty-guided optimization, which brings significant improvement and achieves the state-of-the-art performance on benchmark datasets.
- 169
- Semantics-Aligned Representation Learning for Person Re-Identification164
A framework that drives the reID network to learn semantics-aligned feature representation through delicate supervision designs is proposed and achieves the state-of-the-art performances on the benchmark datasets CUHK03, Market1501, MSMT17, and the partial person reID dataset Partial REID.
- MetaAlign: Coordinating Domain Alignment and Classification for Unsupervised Domain Adaptation135
This paper proposes an effective meta-optimization based strategy dubbed MetaAlign, which treats the domain alignment objective and the classification objective as the meta-train and meta-test tasks in a meta-learning scheme, and encourages both tasks to be optimized in a coordinated way.
- Multi-Granularity Reference-Aided Attentive Feature Aggregation for Video-Based Person Re-Identification128
This paper proposes an attentive feature aggregation module, namely Multi-Granularity Reference-aided Attentive Feature Aggregation (MG-RAFA), to delicately aggregate spatio-temporal features into a discriminative video-level feature representation.
- Uncertainty-Aware Multi-Shot Knowledge Distillation for Image-Based Object Re-Identification127
This paper designs an Uncertainty-aware Multi-shot Teacher-Student (UMTS) Network that consists of a teacher network that learns the comprehensive features from multiple images of the same object, and a student network that takes a single image as input.
- Adaptive Frequency Filters As Efficient Global Token Mixers124
This work applies conventional convolution theorem to deep learning and reveals that adaptive frequency filters can serve as efficient global token mixers, and takes AFF token mixers as primary neural operators to build a lightweight neural network, dubbed AFFNet.
- Style Normalization and Restitution for Domain Generalization and Adaptation113
A novel Style Normalization and Restitution module (SNR) is designed to simultaneously ensure high generalization and discrimination capability of the networks and enforce a dual restitution loss constraint to encourage the better separation of task-relevant and task-irrelevant features.
- Adding Attentiveness to the Neurons in Recurrent Neural Networks102
The effectiveness of the proposed EleAtt-RNN is demonstrated by applying it to the action recognition tasks on both 3D human skeleton data and RGB videos.
- Deep Frequency Filtering for Domain Generalization97
The effectiveness of the proposed Deep Frequency Filtering is demonstrated and it is shown that applying the DFF on a plain baseline out-performs the state-of-the-art methods on different domain generalization tasks, including close-set classification and open-set retrieval.
- Pose-Guided Feature Learning with Knowledge Distillation for Occluded Person Re-Identification87
A network named Pose-Guided Feature Learning with Knowledge Distillation (PGFL-KD), where the pose information is exploited to regularize the learning of semantics aligned features but is discarded in testing to achieve high accuracy while preserving low inference complexity.
- 76
- 75
- Global Distance-Distributions Separation for Unsupervised Person Re-identification75
A global distance-distributions separation (GDS) constraint over the two distributions is introduced to encourage the clear separation of positive and negative samples from a global view and validation of the effectiveness of the GDS constraint in unsupervised ReID networks is validated.
- Human Pose Estimation Using Global and Local Normalization71
A two-stage normalization scheme, human body normalization and limb normalization, is presented to make the distribution of the relative joint locations compact, resulting in easier learning of convolutional spatial models and more accurate pose estimation.
- 65
- Temporal–Spatial Mapping for Action Recognition64
This work introduces a simple yet effective operation, termed temporal–spatial mapping, for capturing the temporal evolution of the frames by jointly analyzing all the frames of a video, and proposes a temporal attention model within a shallow convolutional neural network to efficiently exploit the temporal-spatial dynamics.
- View Invariant 3D Human Pose Estimation62
A view-invariant 3D human pose estimation module to alleviate the effects of viewpoint diversity and shows robustness for different 3D pose base networks and achieves a significant improvement over a powerful baseline on the public3D pose estimation benchmark Human3.6M.
- 62
- ObjFormer: Learning Land-Cover Changes From Paired OSM Data and Optical High-Resolution Imagery via Object-Guided Transformer60
This article pioneers the direct detection of land-cover changes utilizing paired OSM data and optical imagery, thereby expanding the scope of CD tasks and raising a new semi-supervised semantic CD (SCD) task that does not require any manually annotated land-cover labels to train semantic change detectors.
- Lifelong Unsupervised Domain Adaptive Person Re-identification with Coordinated Anti-forgetting and Adaptation53
An effective scheme for this task, dubbed CLUDA-ReID, where the anti-forgetting is harmoniously coordinated with the adaptation, and a meta-based Coordinated Data Replay strategy is proposed to replay old data and update the network with a coordinated optimization direction for both adaptation and memorization.
- AttributeNet: Attribute enhanced vehicle re-identification45
This work introduces AttributeNet (ANet) that jointly extracts identity-relevant features and attribute features and proposes a constraint, named Amelioration Constraint (AC), which encourages the feature after adding attribute features onto the general ReID feature to be more discriminative than the original general ReID feature.
- Skeleton-Based Mutually Assisted Interacted Object Localization and Human Action Recognition42
A joint learning framework for mutually assisted “interacted object localization” and “human action recognition” based on skeleton data is proposed and the temporal consistency of interacted object is explored as constraint to better localize the interacted object with the absence of ground-truth labels.
- 42
- WEDGE: Web-Image Assisted Domain Generalization for Semantic Segmentation39
A WEb-image assisted Domain GEneralization (WEDGE) scheme, which is the first to exploit the diversity of web-crawled images for generalizable semantic segmentation and injects styles of the web- crawled images into training images on-the-fly during training, which enables the network to experience images of diverse styles with reliable labels for effective training.
- Uncertainty-Aware Few-Shot Image Classification38
This work proposes Uncertainty-Aware Few-Shot framework for image classification by modeling uncertainty of the similarities of query-support pairs and performing uncertainty-aware optimization and achieves the state-of-the-art performance.
- 38
- 35
- Shatter and Gather: Learning Referring Image Segmentation with Text Supervision34
A new model is presented that discovers semantic entities in input image and then combines such entities relevant to text query to predict the mask of the referent in order to solve the problem of manual labeling of training data.
- Learning Distortion Invariant Representation for Image Restoration from a Causality Perspective33
This paper is the first to propose a novel training strategy for image restoration from the causality perspective, to improve the generalization ability of DNNs for unknown degradations.
- 29
- Active Token Mixer28
This work proposes an innovative token-mixer, dubbed Active Token Mixer (ATM), to actively incorporate contextual information from other tokens in the global scope into the given query token, taking ATMs as the primary operators and assembling them into a cascade architecture, dubbed ATMNet.
- 27
- 22
- Re-energizing Domain Discriminator with Sample Relabeling for Adversarial Domain Adaptation22
This work proposes an efficient optimization strategy named Re-enforceable Adversarial Domain Adaptation (RADA), which aims to re-energize the domain discriminator during the training by using dynamic domain labels, and relabeling makes the less separable distributions more separable, and thus leads to a more powerful domain classifier w.r.t. the new data distributions.
- 21
- Beyond Triplet Loss: Meta Prototypical N-Tuple Loss for Person Re-identification20
A multi-class classification loss, i.e., N-tuple loss, is introduced to jointly consider multiple ($N$) instances for per-query optimization for person ReID, and achieves the state-of-the-art performance on the benchmark person Re ID datasets.
- 19
- GSemSplat: Generalizable Semantic 3D Gaussian Splatting from Uncalibrated Image Pairs16
GSemSplat is introduced, a framework that learns open-vocabulary semantic representations linked to 3D Gaussians without the need for per-scene optimization, dense image collections or calibration, and employs a dual-feature approach that leverages both region-specific and context-aware semantic features as supervision in the 2D space.
- Slot-VLM: Object-Event Slots for Video-Language Modeling14
This work introduces Slot-VLM, a new framework designed to generate semantically decomposed video tokens, in terms of object-wise and event-wise visual representations, to facilitate LLM inference, which achieves the state-of-the-art performance on video question-answering.
- Versatile Neural Processes for Learning Implicit Neural Representations13
This paper introduces a bottleneck encoder that produces fewer and informative context tokens, relieving the high computational cost while providing high modeling capability, and proposes an efficient NP framework dubbed Versatile Neural Processes (VNP), which largely increases the capability of approximating functions.
- 12
- 12
- Relation-Aware Global Attention.11
This paper proposes an effective Relation-Aware Global Attention module for CNNs to fully exploit the global correlations to infer the attention and demonstrates the general applicability of RGA to vision tasks by applying it to the scene segmentation and image classification tasks resulting in consistent performance improvement.
- 9
- FPCR-Net: Feature pyramidal correlation and residual reconstruction for optical flow estimation9
A semi-supervised Feature Pyramidal Correlation and Residual Reconstruction Network (FPCR-Net) for optical flow estimation from frame pairs and proposes a correlation-warping-normalization (CWN) module to efficiently exploit the correlation dependency.
- 9
- Semantic-Aware Message Broadcasting for Efficient Unsupervised Domain Adaptation8
A novel method, Semantic-aware Message Broadcasting (SAMB), which enables more informative and flexible feature alignment for unsupervised domain adaptation (UDA) and systematically study the effects of adversarial-based feature alignment (ADA) and pseudo-label based self-training (PST) on UDA.
- Vector-based Representation is the Key: A Study on Disentanglement and Compositional Generalization8
This study demonstrates that it is possible to achieve both good concept recognition and novel concept composition, contributing an important step towards human-like intelligence.
- 8
- 7
- 7
- 7
- CASINet: Content-Adaptive Scale Interaction Networks for scene parsing7
A Content-Adaptive Scale Interaction Network (CASINet) to exploit the multi-scale features for scene parsing based on the classic Atrous Spatial Pyramid Pooling module, followed by a proposed contextual scale interaction (CSI) module, and a scale adaptation (SA) module.
- Text Grouping Adapter: Adapting Pre-Trained Text Detector for Layout Analysis6
This paper presents Text Grouping Adapter (TGA), a module that can enable the utilization of various pretrained text detectors to learn layout analysis, allowing us to adopt a well-trained text detector right off the shelf or just fine-tune it efficiently.
- 6
- 5
- Rethinking Classification Loss Designs for Person Re-identification with a Unified View.5
It is argued that triplet-based optimization can be viewed as a two-class subsampling classification, which performs classification over two sampled categories based on instance similarities, and a new meta prototypical N-tuple loss is proposed for more efficient multi- class subsampled classification.
- Exploring the Semantics for Visual Relationship Detection.5
This work proposes a new framework, called semantics guided graph relation neural network (SGRN), for effective visual relationship detection, and introduces a source-target class cognoscitive transformation that transforms the features of the co-occurent objects to the target object domain to refine the visual features.
- TIV-Diffusion: Towards Object-Centric Movement for Text-driven Image to Video Generation4
This work proposes a new diffusion-based TI2V framework, termed TIV-Diffusion, intending to achieve precise control and high-quality video generation based on textual-described motion for different objects, and introduces an object-centric textual-visual alignment module.
- 4
- 4
- 4
- 4
- Deciphering Functions of Neurons in Vision-Language Models3
It is found that there are neurons responsible for only visual or text information, or both, respectively, which are referred to them as visual neurons, text neurons, and multi-modal neurons, respectively.
- 3
- 3
- 2
- 1
- 1
- FPCR-Net: Feature Pyramidal Correlation and Residual Reconstruction for Semi-supervised Optical Flow Estimation1
A semi-supervised Feature Pyramidal Correlation and Residual Reconstruction Network (FPCR-Net) for optical flow estimation from frame pairs and proposes a correlation-warping-normalization (CWN) module to efficiently exploit the correlation dependency.
- –
- –
- Inter Frame Coding with Adaptive Transform–
An adaptive transform scheme to further exploit the non-local correlation for the motion-compensated residual by fully exploring the correlation of abundant similar blocks achieves 0.1~0.5 dB gain in term of PSNR at high bit rate over the state-of-the-art scheme.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.