Title :
Robust speech recognition using cepstral domain missing data techniques and noisy masks
Author_Institution :
ESAT, Katholieke Univ., Leuven, Heverlee, Belgium
Abstract :
Missing data techniques (MDT) have been shown to be an effective method for curing the performance degradation of HMM-based speech recognition systems operating on noisy signals. However, a major drawback of the approach is that MDT requires that the acoustic model be expressed as a mixture of diagonal Gaussians in the log-spectral domain, whereas a higher accuracy can be obtained with Gaussian mixtures in the cepstral domain. The paper describes a recognizer based on the recently described cepstral-domain MDT approach using missing data masks computed from the noisy signal. It exploits a novel decision criterion that integrates harmonicity with signal-to-noise ratio and which makes minimal assumptions on the noise. The system is shown to exhibit a recognition accuracy that is comparable to the ETSI advanced front-end reference.
Keywords :
Gaussian processes; acoustic noise; cepstral analysis; decision making; random noise; speaker recognition; cepstral domain missing data techniques; decision criterion; diagonal Gaussians; harmonicity; log-spectral domain; noisy masks; robust speech recognition; signal-to-noise ratio; Acoustic noise; Cepstral analysis; Curing; Degradation; Gaussian processes; Hidden Markov models; Robustness; Signal to noise ratio; Speech recognition; Telecommunication standards;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 2004. Proceedings. (ICASSP '04). IEEE International Conference on
Print_ISBN :
0-7803-8484-9
DOI :
10.1109/ICASSP.2004.1325960