Fourth European Conference on Speech Communication and Technology

Madrid, Spain
September 18-21, 1995

Speech Enhancement by Eigen Decomposition with Two-Channel Observations

Yuchang Cao, Sridha Sridharan, Miles Moody

Speech Research Lab.,Signal Processing Research Centre, School of Electrical and Electronic Systems Engineering, Queensland University of Technology, Brisbane, Australia

The concept of eigen decomposition for two-channel signal separation, a novel method of enhancing the desired signal corrupted by interference, is described. The method uses two observations, which come from a pair of single sensors (or beamformers) both of which contain the desired signal and the undesired signal(s). The method assumes that the desired signal and the undesired signal(s) are uncorrelated and the signal-to-noise ratios (the ratio of the desired signal to the undesired signal(s)) of each observation are different. A new algorithm based on eigen analysis has been developed to eliminate the noise components sequentially. The technique has been successfully used to separate a speech signal corrupted heavily by ambient noise, co-talker interference and background music.

Full Paper

Bibliographic reference.  Cao, Yuchang / Sridharan, Sridha / Moody, Miles (1995): "Speech enhancement by eigen decomposition with two-channel observations", In EUROSPEECH-1995, 2017-2020.