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

Aalborg, Denmark
September 3-7, 2001

                 

Speech Enhancement and Source Separation based on Binaural Negative Beamforming

A. Álvarez-Marquina, P. Gómez-Vilda, R. Martínez-Olalla, V. Nieto-Lluís, V. Rodellar-Biarge

Universidad Politécnica de Madrid, Spain

Negative Beamformers are well known for their high angular selectivity, which makes them potentially suitable for speech enhancement applications in noisy backgrounds and for directional source separation. On the other hand, Spectral Subtraction is a well-known method for removing noise from a noise-corrupted speech signal. The scheme that is proposed in this paper combines both techniques in order to obtain large gains in the SNR at a reasonable low computational cost. This method may be used to eliminate or enhance a specific signal using a binaural array. The fundamentals of the technique are reviewed, and a structure to control and improve its angular selectivity is presented. Results obtained in a real situation are also commented. Applications of this technique may be found in Security Systems, Domotic Control and also, to improve Speech Recognition.

Full Paper

Bibliographic reference.  Álvarez-Marquina, A. / Gómez-Vilda, P. / Martínez-Olalla, R. / Nieto-Lluís, V. / Rodellar-Biarge, V. (2001): "Speech enhancement and source separation based on binaural negative beamforming", In EUROSPEECH-2001, 2615-2618.