Hongliang Zeng
Papers
1
Total Citations
11
H-Index
1
About
Hongliang Zeng is a researcher advancing reinforcement learning (RL) for robotics, with a focus on enabling agents to learn complex control tasks under sparse reward conditions. His key research areas include robot control, goal-conditioned RL, and curiosity-driven exploration. Zeng’s major contribution is the development of the Adaptive Hindsight Experience Replay with Goal-Amended Curiosity Module (AHEGC), a novel framework that integrates adaptive hindsight replay with an intrinsic curiosity mechanism to shape reward signals autonomously. This approach significantly reduces the need for manually engineered reward functions, a longstanding bottleneck in applying RL to real-world robotics. The AHEGC paper, published in 2023, has already garnered 11 citations, reflecting its timely impact on the field. By addressing the challenge of sparse rewards, Zeng’s work paves the way for more efficient and scalable robot learning systems. His research is particularly notable for its practical orientation, aiming to bridge the gap between algorithmic advances and real-world robotic control.
Research Focus
Key Achievements
Top Papers
- 1