Projects

• CAREER: Compositional Learning and Understanding of the Physical World, IIS-2442540, 6/17/2025-6/30/2030

Code

• RARE: Learn to Rank and Retrieve for Monocular 3D Object Detection (CVPR 2026)
• FairScene: Learning Class-Disentangled 2D/3D Representations for Semantic Scene Completion (WACV 2026)
• Modeling and Learning Multiple Hypotheses for Monocular 3D Object Detection (WACV 2026)
• Learning Partonomic 3D Reconstruction from Image Collections (CVPR 2025)
• Towards Unsupervised Learning of Joint Facial Landmark Detection and Head Pose Estimation (PR 2025)
• Analysis-by-Synthesis Transformer for Single-View 3D Reconstruction (ECCV 2024)
• Learning Anomalies with Normality Prior for Unsupervised Video Anomaly Detectionn (ECCV 2024)
• Imbalance-Aware Discriminative Clustering for Unsupervised Semantic Segmentation (IJCV 2024)
• Towards Generalizable Multi-Object Tracking (CVPR 2024)
• Fully Test-time Adaptation for Object Detection (CVPRW 2024)
• Loss Functions for Pose Guided Person Image Generation (PR 2022)
• Modulated Graph Convolutional Network for 3D Human Pose Estimation (ICCV 2021)
• Enriching Local and Global Contexts for Temporal Action Localization (ICCV 2021)
• Meta Pairwise Relationship Distillation for Unsupervised Person Re-identification (ICCV 2021)
• Addressing Class Imbalance in Scene Graph Parsing by Learning to Contrast and Score (ACCV 2020)
• High-order Graph Convolutional Networks for 3D Human Pose Estimation (BMVC 2020)
• Does Learning Specific Features for Related Parts Help Human Pose Estimation? (CVPR 2019)
• Deeply Learned Compositional Models for Human Pose Estimation (ECCV 2018)
• Towards a Unified Compositional Model for Visual Pattern Modeling (ICCV 2017)
• Sparse Unmixing of Hyperspectral Data Using Spectral a Priori Information (T-GRS 2015)
• Regularized Simultaneous Forward-backward Greedy Algorithm for Sparse Unmixing of Hyperspectral Data (T-GRS 2014)
• Subspace Matching Pursuit for Sparse Unmixing of Hyperspectral Data (T-GRS 2014)
• Nonnegative Matrix Factorization for Hyperspectral Unmixing Using Prior Knowledge of Spectral Signatures (OE 2012)