Zhiyang He
Papers
1
Total Citations
58
H-Index
1
About
Zhiyang He is a leading researcher in embodied AI and visual perception, best known for pioneering simulation-to-reality frameworks that accelerate robotic learning. His landmark work, "Gibson Env: Real-World Perception for Embodied Agents" (2018, 58 citations), introduced a high-fidelity virtual environment that bridges the gap between simulated training and real-world deployment. By enabling agents to learn complex sensorimotor control and active perception tasks in realistic 3D spaces, He’s contributions have fundamentally reduced the cost and fragility barriers of physical robotics. His research integrates computer vision, reinforcement learning, and simulation design, empowering embodied agents to navigate, manipulate, and perceive environments with unprecedented efficiency. He has also advanced scalable benchmarks for evaluating visual navigation and interaction, influencing subsequent work in Habitat and iGibson. With growing citation impact across robotics and AI venues, Zhiyang He’s work continues to shape how autonomous systems learn from virtual worlds to operate robustly in the physical one—making him a key figure in the next generation of intelligent, embodied agents.
Research Focus
Key Achievements
Top Papers
- 1Gibson Env: Real-World Perception for Embodied Agents58 citations · 2018