Jun Yamada
Papers
4
Total Citations
29
H-Index
3
About
Jun Yamada’s research lies at the intersection of robot learning, motion planning, and manipulation, with a focus on enabling robots to operate effectively in cluttered, obstructed environments. His major contributions include pioneering the integration of motion planners with deep reinforcement learning (RL) to overcome the data inefficiency of RL in contact-rich tasks, as demonstrated in his top-cited work (15 citations). He has also advanced gradient-based motion planning in structured latent spaces, leveraging scene embeddings to achieve faster computation without sacrificing planning success. Additionally, Yamada has developed efficient skill acquisition methods for complex manipulation, requiring only a few demonstrations, and introduced a novel approach for learning dual-arm manipulation from human demonstrations translated to a robotic arm. His work addresses critical challenges in small-batch assembly and real-world applicability, with cumulative citations reflecting growing impact. Notably, his research on scene embeddings for latent-space planning represents a significant step toward bridging simulation and reality, making his contributions highly relevant for students and researchers in robotics and AI.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4