Recognition experiment of Korean connected digit telephone speech using the temporal filter based on training speech data

훈련데이터 기반의 temporal filter를 적용한 한국어 4연숫자 전화음성의 인식실험

  • Published : 2003.10.01

Abstract

In this paper, data-driven temporal filter methods[1] are investigated for robust feature extraction. A principal component analysis technique is applied to the time trajectories of feature sequences of training speech data to get appropriate temporal filters. We did recognition experiments on the Korean connected digit telephone speech database released by SITEC, with data-driven temporal filters. Experimental results are discussed with our findings.

Keywords