Yixin Zhu

Yixin Zhu

Assistant Professor

PKU Institute for AI

Biography

Dr. Yixin Zhu received a Ph.D. degree (‘18) from UCLA advised by Prof. Song-Chun Zhu. His research builds interactive AI by integrating high-level common sense (functionality, affordance, physics, causality, intent) with raw sensory inputs (pixels and haptic signals) to enable richer representation and cognitive reasoning on objects, scenes, shapes, numbers, and agents. Dr. Zhu directs the PKU CoRe Lab, working on abstract reasoning, visually grounded reasoning, and interactive reasoning.

[CV] [Dissertation]

Getting Started 帮你自学大学四年AI的所有内容
通班相关文章 汇集官方和权威渠道对通班的报道
博雅一号(成都) 目前仅限AI院全职PI和北大通班学生使用

Pre-Print

GlobalTomo: A global dataset for physics-ML seismic wavefield modeling and FWI
PhyRecon: Physically Plausible Neural Scene Reconstruction
Dexterous Functional Pre-Grasp Manipulation with Diffusion Policy
Ag2Manip: Learning Novel Manipulation Skills with Agent-Agnostic Visual and Action Representations
PreAfford: Universal Affordance-Based Pre-Grasping for Diverse Objects and Environments
Zero-Shot Image Feature Consensus with Deep Functional Maps
Tac-Man: Tactile-Informed Prior-Free Manipulation of Articulated Objects

Publications

[CogSci24] Evaluating and Modeling Social Intelligence: A Comparative Study of Human and AI Capabilities
[CVPR24] Move as You Say, Interact as You Can: Language-guided Human Motion Generation with Scene Affordance
[CVPR24] Scaling Up Dynamic Human-Scene Interaction Modeling
[RA-L24] Grasp Multiple Objects with One Hand
[ICLR24] Neural-Symbolic Recursive Machine for Systematic Generalization
[NeurIPS23] ChimpACT: A Longitudinal Dataset for Understanding Chimpanzee Behaviors
[NeurIPS23] Evaluating and Inducing Personality in Pre-trained Language Models
[IROS23] Learning a Causal Transition Model for Object Cutting
[IROS23] Part-level Scene Reconstruction Affords Robot Interaction
[IROS23] Sequential Manipulation Planning for Over-actuated Unmanned Aerial Manipulators