Kyoto University/Honda Research Institute Japan Co.,Ltd.
会議概要（会議名, 開催地, 会期, 主催者等）
LREC 2018 Special Speech Sessions "Speech Resources Collection in Real-World Situations"; Phoenix Seagaia Conference Center, Miyazaki; 2018-05-09
抄録(英)
We have developed the MPR multimodal dialogue corpus and describe research activities using the corpus aimed for enabling multiparty human-robot verbal communication in real-world settings. While aiming for that as the final goal, the immediate focus of our project and the corpus is non-verbal communication, especially social signal processing by machines as the foundation of human-machine verbal communication. The MPR corpus stores annotated audio-visual recordings of dialogues between one robot and one or multiple (up to tree) participants. The annotations include speech segment, addressee of speech, transcript, interaction state, and, dialogue act types. Our research on multiparty dialogue management, boredom recognition, response obligation recognition, surprise detection and repair detection using the corpus is briefly introduced, and an analysis on repair in multiuser situations is presented. It exhibits richer repair behaviors and demands more sophisticated repair handling by machines.
雑誌名
Proceedings of the LREC 2018 Special Speech Sessions
ページ
35 - 39
発行年
2018-05-09
フォーマット
application/pdf
著者版フラグ
publisher
出版者
Center for Corpus Development, National Institute for Japanese Language and Linguistics
A multimodal multiparty human-robot dialogue corpus for real world interaction
lrec2018sss_funakoshi.pdf
(1.53MB)
運用指針、オープンアクセス方針、等
