Joey Tianyi Zhou

3 papers CFAR, A*STAR

• Beyond Modality Collapse: Representation Blending for Multimodal Dataset Distillation
• PandaPose: 3D Human Pose Lifting from a Single Image via Propagating 2D Pose Prior to 3D Anchor Space
• SCOPE: Saliency-Coverage Oriented Token Pruning for Efficient Multimodel LLMs