International Workshop on Hands-Free Speech Communication (HSC2001) April 9-11, 2001 Kyoto, Japan

Speech Acquisition Through Space Diversity Using Distributed Multi-Microphone

Shoji Kajita, Yasuhiro Shimizu, Kazuya Takeda and Fumitada Itakura

Center for Integrated Acoustic Information Research (CIAW), Nagoya University, Japan

This paper proposes space diversity speech recognition technique using distributed multi-microphone in room, as a new paradigm of speech recognition. The key technology to realize the system is (1) distant-talking speech recognition and (2) the integration method of militiple inputs. In this paper, we propose the use of distant speech model for the distant-talking speech recognition, and feature-based and likelihood-based integration methods for multi-microphone distributed in room. The distant speech model is a set of HMMs learned using speech data convolved with the impulse responses measured at several points in room. The experimental results of simulated distant-talking speech recognition show that the proposed space diversity speech recognition system can attain about 80% in accuracy, while the performances of conventional HMM using close-talking microphone are less than 50%. These results indicate that the space diversity approach is promising for robust speech recognition under the real acoustic environment.

Full Paper

Bibliographic reference.  Kajita, Shoji / Shimizu, Yasuhiro / Takeda, Kazuya / Itakura, Fumitada (2001): "Speech acquisition through space diversity using distributed multi-microphone", In HSC2001, 63-66.