Acoustic feature transformation is widely used to reduce dimensionality and improve speech recognition performance. In this letter we focus on dimensionality reduction methods that minimize the average classification error. Unfortunately, minimization of the average classification error may cause considerable overlaps between distributions of some classes. To mitigate risks of considerable over...

In this paper, we first introduce the use of Gaussian mixture models (GMM) for Chinese tone classification in continuous speech. Then, we explain how to integrate it with the HMM-based speech recognition system. Finally, we provide the tone classification accuracy of this probabilistic method which is tested with Chinese continuous speech database of national “863” project.

Speech is an integral part of communication. Speech disorder is a problem with fluency, voice, and or how a person produces a speech sound. The main focus of this study is to identify the difference between normal and disordered speech. The proposed work classifies the normal and abnormal speech. The experimental investigation elucidated MFCC and DTW with the accuracy rate of 88 % and 75% respe...

We explore a speech-based topic classification approach. We generate the transcript of input video lecture based on speech recognition technology and identify the topic by comparing its term-based vector with topic models. The preliminary experiment result shows that the speech-based topic classification works well, with its performance comparable to one that directly uses manual transcripts. T...