INTERSPEECH 2004 - ICSLP 8th International Conference on Spoken Language Processing Jeju Island, Korea October 4-8, 2004

Compact Acoustic Model for Embedded Implementation

Junho Park, Hanseok Ko

Korea University, Seoul, Korea, Korea

An acoustic model for an embedded speech recognition system must exhibit two desirable features; ability to minimize performance degradation in recognition while solving the memory problem under limited system resources. To cope with the challenges, we introduce the state-clustered tied-mixture (SCTM) HMM as an acoustic model optimization. The proposed SCTM modeling shows a significant improvement in recognition performance as well as a solution to sparse training data problem. Moreover, the state weight quantizing method achieves a drastic reduction in model size. In this paper, we describe the acoustic model optimization procedure for embedded speech recognition system and corresponding performance evaluation results.

Full Paper

Bibliographic reference.  Park, Junho / Ko, Hanseok (2004): "Compact acoustic model for embedded implementation", In INTERSPEECH-2004, 693-696.