Papers
2
Total Citations
236
H-Index
2
About
Po-Wei Chou is a leading researcher at the intersection of robotics, simulation, and reinforcement learning, with a focus on enabling more realistic and data-efficient physical interaction for autonomous systems. His work is defined by two major contributions: the development of high-fidelity tactile simulation and the advancement of policy gradient methods for continuous control. Chou is the lead creator of **TACTO**, a fast, flexible, and open-source simulator for high-resolution vision-based tactile sensors. This tool has become an essential resource for the robotics community, enabling researchers to prototype and benchmark tactile perception algorithms without requiring expensive physical hardware—a contribution that has earned **over 130 citations** since its 2022 release. Prior to this, Chou made a foundational impact on deep reinforcement learning for real-world robotics. His 2017 paper on **improving stochastic policy gradients using the Beta distribution** addressed a critical limitation in continuous control: the physical constraints on action spaces. By replacing the Gaussian distribution with a bounded Beta distribution, his method significantly improved policy performance and stability, a breakthrough that has been cited **more than 100 times**. Through these innovations, Chou has helped bridge the gap between simulation and reality, empowering robots to learn more robust and physically grounded manipulation skills.
Research Focus
Key Achievements
Top Papers
- 1
- 2