معرفی
Tetsunori Kobayashi is a Professor in the School of Fundamental Science and Engineering at Waseda University, Japan, where he has served since 1997. He is renowned for pioneering research in human–robot interaction, spoken language processing, and multimodal conversational systems, leading to over 230 refereed papers and an h-index of 35 (Google Scholar).
Education:
- 1980 B.Eng. in Electrical Engineering, Waseda University
- 1982 M.Eng. and 1985 Dr.Eng. from Graduate School of Science and Engineering, Waseda University
Research Interests:
His work spans intelligent robotics, perceptual information processing, pattern recognition, image and audio processing, and conversational AI. He develops algorithms for real-time dialogue systems, multi-party conversation facilitation robots, and non-autoregressive speech recognition leveraging CTC and pre-trained language models.
Recent Publication Trends:
Since 2020 his group has advanced non-autoregressive end-to-end ASR (Mask-CTC, Intermpl, BECTRA), noise-robust attention, multi-look-ahead conversational ASR, and neural speaker diarization. They integrate BERT-style pre-training with CTC losses to accelerate inference while maintaining accuracy. Parallel work explores vision-and-language topics such as scene-graph generation, video semantic indexing, and personalized summarization for spoken news delivery.
Scientific Awards:
- IEICE Fellow 2023 – for multi-modal multi-party conversation research
- IPSJ Fellow 2016 – for pioneering robot conversation studies
- JST Award for Academic Start-ups 2024
- Best Paper Awards from IEICE, IEEE BTAS, ACM SIGGRAPH VRCAI, and several IPSJ workshop prizes
Advising & Grants:
He has mentored dozens of PhD and Master’s students who now lead in academia and industry. Major funded projects include JST CREST on conversational robotics, NEDO and JST-support for AI-based speech interfaces, and industry collaborations with NHK, OKI, and NEC.
Labs & Teams:
Kobayashi heads the Perceptual Computing Laboratory at Waseda, conducting interdisciplinary research with domestic and international partners such as MIT, ATR, and NHK Science & Technology Labs.