Papers
13
Total Citations
214
H-Index
9
About
Hisashi Kawai is a prominent researcher specializing in human-robot interaction, multimodal language understanding, and natural language processing for domestic service robots. His work centers on enabling robots to interpret ambiguous, unconstrained natural language instructions—a critical challenge for deploying robots in real-world home environments. Kawai's most significant contributions lie in developing sophisticated multimodal frameworks that bridge language and visual perception. His GAN-based multimodal target-source classification systems, which garnered over 41 citations, allow robots to correctly identify objects and their locations from everyday fetching instructions. Complementary work on attention branch networks and transformer-based approaches—including the CrossMap Transformer for vision-and-language navigation—further demonstrates his commitment to robust, context-aware robot communication. Beyond instruction understanding, Kawai has made notable strides in robot speech naturalness, pioneering HMM-based, non-monologue speech synthesis through cloud robotics architectures, addressing the longstanding problem of robotic voices sounding monotonous and unengaging. His earlier research on situated spoken dialogue with active learning also tackled the practical safety risks of robot misunderstanding. With over 190 cumulative citations across his key publications, Kawai's research meaningfully advances the field of communicative domestic service robots, making him an important figure for students exploring human-robot interaction and embodied AI.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3A cloud robotics approach towards dialogue-oriented robot speech18 citations · 2015
- 4
- 5Multimodal Attention Branch Network for Perspective-Free Sentence Generation17 citations · 2019
- 6Situated Spoken Dialogue with Robots Using Active Learning16 citations · 2011
- 7
- 8
- 9
- 10