Papers
31
Total Citations
527
H-Index
13
About
Sungjoon Choi is a robotics and artificial intelligence researcher whose work spans motion synthesis, human-robot interaction, and robot navigation. His research bridges the gap between intelligent machine behavior and natural human communication, with particular emphasis on enabling robots to understand and respond to language-based instructions and human motion cues. Choi's most influential contribution is FLAME (2023, 121 citations), a diffusion-based model for free-form language-driven motion synthesis and editing, which has rapidly become a landmark work in the growing field of text-to-motion generation with applications in animation, gaming, and robotics. His earlier work on Gaussian process motion control (2014, 42 citations) demonstrated sophisticated real-time navigation in crowded dynamic environments, while complementary research on vision-based coordinated localization (2014, 37 citations) addressed autonomous operation in GPS-denied settings. Choi has also made notable contributions to humanoid robotics, including motion retargeting pipelines and lifelike robot gaze systems, as well as natural language-driven human-robot collaboration through his IT2P and CLARA frameworks. His work on Tsallis entropy reinforcement learning further reflects a breadth extending into theoretical machine learning. Collectively, his research has accumulated over 385 citations, establishing him as a versatile and impactful voice in intelligent robotics.
Research Focus
Key Achievements
Top Papers
- 1FLAME: Free-Form Language-Based Motion Synthesis & Editing121 citations · 2023
- 2
- 3Vision-Based Coordinated Localization for Mobile Sensor Networks37 citations · 2014
- 4
- 5
- 6Realistic and Interactive Robot Gaze30 citations · 2020
- 7
- 8Nonparametric Motion Retargeting for Humanoid Robots on Shared Latent Space21 citations · 2020
- 9
- 10