TY - GEN
T1 - Dual-microphone noise reduction based on semi-blind DUET
AU - Fu, Zhong Hua
AU - Xie, Lei
AU - Jiang, Domg Mei
PY - 2010
Y1 - 2010
N2 - Noise reduction is a fundamental requirement of many speech applications. Sometimes the major interferences, such as music, song, cross-talking, etc., coming from loudspeakers nearby make it a more challenging problem. In this paper, by supposing the locations of interferences are fixed, a semi-blind Degenerate Unmixing Estimation Technique (DUET) approach using dualmicrophone is proposed. Firstly, during system initialization, the spatial feature distribution of the interferences is estimated precisely. Then the feature distribution of unknown target speech is obtained using model adaptation. Finally, a time-frequency binary mask estimated based on likelihood comparison is used to separate the target speech. The experimental results show that the performances of our approach in adverse noise reduction are significant. 77.9% interferences are eliminated and the speech quality is preserved very well. The results are close to those of un-blind approach, where the locations of all sound sources are known.
AB - Noise reduction is a fundamental requirement of many speech applications. Sometimes the major interferences, such as music, song, cross-talking, etc., coming from loudspeakers nearby make it a more challenging problem. In this paper, by supposing the locations of interferences are fixed, a semi-blind Degenerate Unmixing Estimation Technique (DUET) approach using dualmicrophone is proposed. Firstly, during system initialization, the spatial feature distribution of the interferences is estimated precisely. Then the feature distribution of unknown target speech is obtained using model adaptation. Finally, a time-frequency binary mask estimated based on likelihood comparison is used to separate the target speech. The experimental results show that the performances of our approach in adverse noise reduction are significant. 77.9% interferences are eliminated and the speech quality is preserved very well. The results are close to those of un-blind approach, where the locations of all sound sources are known.
KW - Dual-microphone
KW - DUET
KW - Noise reduction
KW - Semiblind
UR - http://www.scopus.com/inward/record.url?scp=79955878161&partnerID=8YFLogxK
U2 - 10.1109/ISCSLP.2010.5684881
DO - 10.1109/ISCSLP.2010.5684881
M3 - 会议稿件
AN - SCOPUS:79955878161
SN - 9781424462469
T3 - 2010 7th International Symposium on Chinese Spoken Language Processing, ISCSLP 2010 - Proceedings
SP - 33
EP - 37
BT - 2010 7th International Symposium on Chinese Spoken Language Processing, ISCSLP 2010 - Proceedings
T2 - 2010 7th International Symposium on Chinese Spoken Language Processing, ISCSLP 2010
Y2 - 29 November 2010 through 3 December 2010
ER -