81001954

open access

Version of Record

オンセイ キョウチョウ シュホウ オモチイタ コウオン ショウガイ シャ ノ オンセイ ニンシキ

Dysarthric speech recognition using speech enhancement

Japanese (日本語)

神戸大学都市安全研究センター研究報告

13

75-80

神戸大学都市安全研究センター

2009-03

2010-03-19

Recently, the accuracy of speaker-independent speechr ecognition has been remarkably improved by use of stochastic modeling of speech. However there has been very little research on orally-challenged people , such as those with speech impediments. In this paper, we discuss our efforts to build an acoustic model for a person with articulation disorders. The speaking style of a person with an articulation disorder differs cons iderably from that of a physically unimpaired person, and we need a feature extraction characteristic of a person with an articulation disorder. The sound spoken by a dysarthric speaker is less clear than a physically unimpaired person, and spectral information is buried in the sound. Therefore, we investigate a feature extraction method based on LPE (Local Peak Enhanc ement).

紀要論文