Area of research
Computer Vision and Pattern Recognition · Artificial Intelligence
Research interest
Research interests include Advanced Neural Network Applications, Domain Adaptation and Few-Shot Learning, Video Surveillance and Tracking Methods, and Advanced Image and Video Retrieval Techniques.
A Task-Aware Parameter Decoupling Framework for Continual Anomaly Detection
From sparse semantics to rich instances: Empowering label-efficient LiDAR panoptic segmentation via geometric priors.
FocusPatch AD: Few-Shot Multi-Class Anomaly Detection With Unified Keywords Patch Prompts.
Transporting the Cross-Modal Prototypes for Unsupervised Visible-Infrared Person Re-Identification.
Decoupling 3-D Point Cloud Attributes for Semantic Segmentation via Real-World Prior Exploitation
Recent Advances of Foundation Language Models-based Continual Learning: A Survey
GEOcc: Geometrically Enhanced 3D Occupancy Network With Implicit-Explicit Depth Fusion and Contextual Self-Supervision
Adaptive Pseudo-Label Purification and Debiasing for Unsupervised Visible-Infrared Person Re-Identification
Cross-Modal Recipe Retrieval With Fine-Grained Prompting Alignment and Evidential Semantic Consistency
Fusion-Then-Distillation: Toward Cross-Modal Positive Distillation for Domain Adaptive 3D Semantic Segmentation
RE-GZSL: Relation Extrapolation for Generalized Zero-Shot Learning
SeqVLM: Proposal-Guided Multi-View Sequences Reasoning via VLM for Zero-Shot 3D Visual Grounding
REDAD: A Reliable Distillation for Image Anomaly Detection
Gradient Projection for Continual Parameter-Efficient Tuning.
Bias to Balance: New-Knowledge-Preferred Few-Shot Class-Incremental Learning via Transition Calibration.
CloudMix: Dual Mixup Consistency for Unpaired Point Cloud Completion.
PLATO-TTA: Prototype-Guided Pseudo-Labeling and Adaptive Tuning for Multi-Modal Test-Time Adaptation of 3D Segmentation
PFDepth: Heterogeneous Pinhole-Fisheye Joint Depth Estimation via Distortion-aware Gaussian-Splatted Volumetric Fusion
MOL: Joint Estimation of Micro-Expression, Optical Flow, and Landmark via Transformer-Graph-Style Convolution.
Cross-Cloud Consistency for Weakly Supervised Point Cloud Semantic Segmentation.
Novel Category Discovery with X-Agent Attention for Open-Vocabulary Semantic Segmentation
Glass Makes Blurs: Learning the Visual Blurriness for Glass Surface Detection
Multi-memory Matching for Unsupervised Visible-Infrared Person Re-identification
Robust Pseudo-label Learning with Neighbor Relation for Unsupervised Visible-Infrared Person Re-Identification
CLIP2UDA: Making Frozen CLIP Reward Unsupervised Domain Adaptation in 3D Semantic Segmentation
EdgeFormer: Edge-Aware Efficient Transformer for Image Super-Resolution
Semisupervised Progressive Representation Learning for Deep Multiview Clustering.
Learning All-In Collaborative Multiview Binary Representation for Clustering.
PIG: Prompt Images Guidance for Night-Time Scene Parsing.
Variational Distillation for Multi-View Learning.