5th European Conference on Speech Communication and Technology

Rhodes, Greece
September 22-25, 1997

Designing a Reduced Feature-Vector Set for Speech Recognition by Using KL/GPD Competitive Training

Tsuneo Nitta, Akinori Kawamura

Multimedia Engineering Laboratory, Toshiba Corporation, Kawasaki, Japan

The hybrid algorithm of SMQ (Statistical Matrix Quantization) and HMM shows high performance in vocabulary-unspecific, speaker-independent speech recognition, however, it needs lots of computation and memory at the stage of the segment quantizer of SMQ. In this paper, we propose a newly developed, two-stage segment quantizer with a feature extractor based on KL expansion and a classifier, that can be trained by using competitive training of KL/GPD. Result of experiments shows 1/30 - 1/40 reduction in both computation time and a memory size with the same performance that the old version of SMQ shows.

Full Paper

Bibliographic reference.  Nitta, Tsuneo / Kawamura, Akinori (1997): "Designing a reduced feature-vector set for speech recognition by using KL/GPD competitive training", In EUROSPEECH-1997, 2107-2110.