Hierarchical and parallel processing of modulation spectrum for ASR applications

• Published in: 2008 IEEE International Conference on Acoustics, Speech and Signal Processing pp. 4165 - 4168
• Main Authors: Valente, F., Hermansky, H.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.03.2008
• Subjects: Automatic speech recognition; Band pass filters; Data mining; Filtering; Fourier transforms; Frequency modulation; Gabor filters; Hierarchical and parallel combination; LUCSR; Modulation spectrum; Multi-resolution filter; Neural Network; Neural networks; Parallel processing; Speech recognition
• ISBN: 9781424414833; 1424414830
• ISSN: 1520-6149; 2379-190X
• DOI: 10.1109/ICASSP.2008.4518572

The modulation spectrum is an efficient representation for describing dynamic information in signals. In this work we investigate how to exploit different elements of the modulation spectrum for extraction of information in automatic recognition of speech (ASR). Parallel and hierarchical (sequential) approaches are investigated. Parallel processing combines outputs of independent classifiers applied to different modulation frequency channels. Hierarchical processing uses different modulation frequency channels sequentially. Experiments are run on a LVCSR task for meetings transcription and results are reported on the RT05 evaluation data. Processing modulation frequencies channels with different classifiers provides a consistent reduction in WER (2% absolute w.r.t. PLP baseline). Hierarchical processing outperforms parallel processing. The largest WER reduction is obtained through sequential processing moving from high to low modulation frequencies. This model is consistent with several perceptual and physiological studies on auditory processing.