Song-Fang Han
Papers
1
Total Citations
2
H-Index
1
About
Song-Fang Han is a leading researcher at the intersection of computer vision, graphics, and robotics, with a core focus on bridging the sim-to-real gap for autonomous systems. Their most notable contribution is the development of physically grounded sensor simulation, exemplified by their pioneering work on active stereo sensors. By designing a fully physics-grounded simulation pipeline—from material acquisition to ray-tracing—Han’s research enables virtual depth sensors to produce outputs that are nearly indistinguishable from real-world data. This breakthrough directly addresses the long-standing challenge of domain shift, allowing perception models trained entirely in simulation to transfer seamlessly to physical robots. While their highly cited work is still emerging, Han’s approach has already garnered attention for its potential to revolutionize how autonomous vehicles and robotic manipulators are trained, drastically reducing the need for expensive real-world data collection. Their research promises to accelerate the deployment of reliable, perception-driven systems in unstructured environments, marking Han as a rising star in embodied AI and photorealistic simulation.
Research Focus
Key Achievements
Top Papers
- 1