In the context of an interactive interface with a computer, a method is proposed for simple representation of dialog states by switching between facial pictures according to voice information. First, essences that represent dialog states are drawn out, and examined separately. Continuous mouth movement during the process of speech is presented as the vocal essence. The importance of the speaker's and listener's movements is shown, and necessity of imitating those movements is explained; this is treated as the realistic essence. As to the formal essence, an algorithm is proposed to represent an interlocutor by a facial picture while bringing its mouth into motion according to loudness. The effectiveness of the proposed method is verified by comparative experiments. The optimal display rate for the proposed method is determined experimentally.
|ジャーナル||Systems and Computers in Japan|
|出版物ステータス||出版済み - 6 30 1998|
All Science Journal Classification (ASJC) codes
- Theoretical Computer Science
- Information Systems
- Hardware and Architecture
- Computational Theory and Mathematics