Min Jae Song
Papers
1
Total Citations
15
H-Index
1
About
Min Jae Song is a robotics researcher whose work lies at the intersection of computer vision, motion planning, and human-robot interaction. His primary research areas include self-supervised learning for motion generation, safe robot control, and humanoid locomotion. Song’s most notable contribution is the development of **Self-Supervised Shared Latent Embedding (S³LE)**, a data-driven motion retargeting method that allows humanoid robots to naturally mimic human motions captured from RGB video or motion capture data—without requiring paired human-robot demonstrations. This work, published in 2021, has garnered **15 citations** and addresses a critical challenge in robotics: generating safe, physically plausible motions while preserving the style and intent of the original human movement. By embedding safety guarantees directly into the learning pipeline, Song’s approach bridges the gap between expressive motion generation and real-world deployability. His research is particularly impactful for applications in assistive robotics, teleoperation, and humanoid animation. With a focus on both theoretical rigor and practical safety, Song is advancing the frontier of robots that can learn and move like humans, making human-robot collaboration more intuitive and trustworthy.
Research Focus
Key Achievements
Top Papers
- 1Self-Supervised Motion Retargeting with Safety Guarantee15 citations · 2021