Gangshan Wu

32

Papers

66

Total Citations

Papers (32)

CATANet: Efficient Content-Aware Token Aggregation for Lightweight Image Super-Resolution

Sketch and Refine: Towards Fast and Accurate Lane Detection

Asymmetric Masked Distillation for Pre-Training Small Foundation Models

Tra-MoE: Learning Trajectory Prediction Model from Multiple Domains for Adaptive Policy Conditioning

SportsHHI: A Dataset for Human-Human Interaction Detection in Sports Videos

Learning Actor Relation Graphs for Group Activity Recognition

Translate-to-Recognize Networks for RGB-D Scene Recognition

Residual Feature Aggregation Network for Image Super-Resolution

TDN: Temporal Difference Networks for Efficient Action Recognition

CGA-Net: Category Guided Aggregation for Point Cloud Semantic Segmentation

MixFormer: End-to-End Tracking With Iterative Mixed Attention

Extracting Motion and Appearance via Inter-Frame Attention for Efficient Video Frame Interpolation

STMixer: A One-Stage Sparse Action Detector

StereoSnakes: Contour Based Consistent Object Extraction For Stereo Images

LIP: Local Importance-Based Pooling

Target Adaptive Context Aggregation for Video Scene Graph Generation

Mutual Supervision for Dense Object Detection

MGSampler: An Explainable Sampling Strategy for Video Action Recognition

Self Supervision to Distillation for Long-Tailed Visual Recognition

MultiSports: A Multi-Person Video Dataset of Spatio-Temporally Localized Sports Actions

Relaxed Transformer Decoders for Direct Action Proposal Generation

Robust Object Modeling for Visual Tracking

SportsMOT: A Large Multi-Object Tracking Dataset in Multiple Sports Scenes

Efficient Video Action Detection with Token Dropout and Context Refinement

Actions as Moving Points

Boundary-Aware Cascade Networks for Temporal Action Segmentation

Context-Aware RCNN: A Baseline for Action Detection in Videos

LinK: Linear Kernel for LiDAR-Based 3D Perception

AutoLUT: LUT-Based Image Super-Resolution with Automatic Sampling and Adaptive Residual Learning

Text-Guided Nonverbal Enhancement Based on Modality-Invariant and -Specific Representations for Video Speaking Style Recognition

Dual DETRs for Multi-Label Temporal Action Detection

MixFormerV2: Efficient Fully Transformer Tracking