Papers
9
Total Citations
164
H-Index
6
About
Daichi Mochihashi is a leading researcher at the intersection of machine learning, robotics, and natural language processing. His work centers on the fundamental challenge of unsupervised segmentation—how machines can autonomously divide continuous, high-dimensional data streams into meaningful units, much like humans segment speech into words or motions into actions. His most influential contributions include the development of probabilistic models that combine Hidden semi-Markov Models (HSMM) with Gaussian Processes (GP-HSMM) to segment continuous motions, a method that has garnered 56 citations. He also pioneered HVGH, a technique using deep neural compression and statistical generative models for high-dimensional time series segmentation. Mochihashi has significantly advanced the field of language and robotics, co-authoring a widely-cited survey on the topic (53 citations) and serving as an editorial voice for the discipline. His work is crucial for enabling robots to learn language and motion from raw sensory input without human annotation, laying the groundwork for more autonomous and adaptable robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2Survey on frontiers of language and robotics53 citations · 2019
- 3
- 4
- 5
- 6
- 7
- 8Robotics, Grounding and Natural Language Processing2 citations · 2020
- 9Editorial: Language and Robotics2 citations · 2021