Area of research
Computer Vision and Pattern Recognition · Computational Mechanics
Research interest
Research interests include 3D Shape Modeling and Analysis, Robotics and Sensor-Based Localization, Advanced Vision and Imaging, and 3D Surveying and Cultural Heritage.
Motion and Appearance Decoupling Representation for Event Cameras
MAC: Maximal Cliques for 3D Registration
Discriminative Correspondence Estimation for Unsupervised RGB-D Point Cloud Registration
Adaptive Sparse Memory Networks for Efficient and Robust Video Object Segmentation
Learning Hierarchical Color Guidance for Depth Map Super-Resolution
GeoTransformer: Fast and Robust Point Cloud Registration With Geometric Transformer
MTU-Net: Multilevel TransUNet for Space-Based Infrared Tiny Ship Detection
RoReg: Pairwise Point Cloud Registration With Oriented Descriptors and Local Rotations
Context-aware Alignment and Mutual Masking for 3D-Language Pre-training
Exploring Hierarchical Spatial Layout Cues for 3D Point Cloud Based Scene Graph Prediction
AnchorPoint: Query Design for Transformer-Based 3D Object Detection and Tracking
2D3D-MATR: 2D-3D Matching Transformer for Detection-free Registration between Images and Point Clouds
Unified Spatio-Temporal Dynamic Routing for Efficient Video Object Segmentation
Geometric Transformer for Fast and Robust Point Cloud Registration
CVCNet: Learning Cost Volume Compression for Efficient Stereo Matching