Sungtak Kim
Papers
2
Total Citations
28
H-Index
2
About
Sungtak Kim is a leading researcher in human-robot interaction, specializing in robust speaker identification for autonomous systems. His work focuses on overcoming the critical challenge of distant-talking speech recognition in noisy, real-world environments, particularly for home and ubiquitous robot companions. Kim's major contribution is the development of a text-independent speaker identification system that leverages multiple microphones and a novel "soft channel selection" technique. This approach intelligently weights and combines identification results from different microphones based on their reliability, dramatically improving accuracy compared to single-microphone systems. His seminal 2008 paper on this method, published in a top venue, has garnered 20 citations, establishing a foundational approach for reliable speaker recognition on mobile robots. His earlier 2007 work, with 8 citations, laid the groundwork for this multi-microphone framework in ubiquitous robot companions. By enabling robots to accurately identify users from a distance without requiring specific phrases, Kim's research directly enhances the naturalness and security of human-robot interaction, paving the way for more intuitive and responsive robotic assistants in homes and public spaces.
Research Focus
Key Achievements
Top Papers
- 1
- 2