Speech enhancement using combination of dereverberation and noise reduction for robust speech recognition

Authors:
Tien Dung Tran;Dang Khoa Nguyen;Thi Anh Xuan Tran;Quoc Cuong Nguyen;Huu Binh Nguyen
Affiliations:
Hanoi University of Science and Technology;Hanoi University of Science and Technology;Hanoi University of Science and Technology;Hanoi University of Science and Technology;Hanoi University of Science and Technology
Venue:
Proceedings of the Second Symposium on Information and Communication Technology
Year:
2011

Citing 6
Cited 1

A Binaural Speech Processing Method Using Subband-Crosscorrelation Analysis for Noise Robust Recognition

ICASSP '97 Proceedings of the 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP '97)-Volume 2 - Volume 2
Spatial separation of speech signals using amplitude estimation based on interaural comparisons of zero-crossings

Speech Communication
Robust speech dereverberation based on non-negativity and sparse nature of speech spectrograms

ICASSP '09 Proceedings of the 2009 IEEE International Conference on Acoustics, Speech and Signal Processing
Robust speech recognition by integrating speech separation and hypothesis testing

Speech Communication
Mask estimation for missing data speech recognition based on statistics of binaural interaction

IEEE Transactions on Audio, Speech, and Language Processing
Phase-based dual-microphone robust speech enhancement

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

Robust speech recognition based on binaural speech enhancement system as a preprocessing step

Proceedings of the Third Symposium on Information and Communication Technology

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we describe a speech enhancement approach for robust speech recognition. This approach consists of two stages to solve both current problems of speech recognition: reverberation and noise. Firstly, speech signal is dereveberated by suppression of slowly -- varying components and the falling edge of the power envelope (SSF). Then a binaural speech processing is applied to remove noise from target speech. Speech recognition results show that this combination algorithm provides a good robustness in real environments.