EUROSPEECH 2001 Scandinavia
7th European Conference on Speech Communication and Technology
2nd INTERSPEECH Event

Aalborg, Denmark
September 3-7, 2001

                  

Using the Modulation Complex Wavelet Transform for Feature Extraction in Automatic Speech Recognition

Yasunori Momomura (1), Kenji Okada (1), Takayuki Arai (1), Noboru Kanedera (2), Yuji Murahara (1)

(1) Sophia Univ., Japan
(2) Ishikawa National College of Technology, Japan

In this paper we examine robust feature extraction methods for automatic speech recognition (ASR) in noise-distorted environments. Previous research showed that combining the coefficients of multi-resolutional modulation frequency band. We show that this multi-resolutional approach can be achieved using a wavelet transform instead of the Fourier transform. Taking the FFT phase into consideration, we applied the Gabor function, which is a complex function, as mother wavelet. This approach yielded a 1.7% increase in recognition accuracy compared to the FFT-based multi-resolutional approach.

Full Paper

Bibliographic reference.  Momomura, Yasunori / Okada, Kenji / Arai, Takayuki / Kanedera, Noboru / Murahara, Yuji (2001): "Using the modulation complex wavelet transform for feature extraction in automatic speech recognition", In EUROSPEECH-2001, 2639-2642.