KMID : 0917520000070040049
|
|
Journal of Speech Sciences 2000 Volume.7 No. 4 p.49 ~ p.57
|
|
Model based Stress Decision Method
|
|
|
|
Abstract
|
|
|
This paper proposes an effective decision method focused on evaluating the "stress position". Conventional methods usually extract the acoustic parameters and compare them to references in absolute scale, adversely producing unstable results as testing conditions change. To code with environmental dependency, the proposed method is designed to be model-based and determines the stressed interval by making relative comparison over candidates. The stressed/unstressed models are then induced from normal phone models by aaptive training. The experimentl results indicate that the proposed method is promising, and that it is useful for automatic detection of stress position. The results also show that generating the stressed/unstressed model by adaptive training is effective.
Keywords: Stress, model-based, adaptive training, stressed/unstressed model, speech recognition
|
|
KEYWORD
|
|
|
|
FullTexts / Linksout information
|
|
|
|
Listed journal information
|
|
|
|