Voice activity classification using beamformer-output-ratio

Tran, Thuy N; Cowley, William G.; Pollok, Andre

doi:10.1109/ausctw.2012.6164913

Cited by 3 publications

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is observed that the LogBPR is positive in the state of S1, the LogBPR is negative in the state of S2, and the LogBPR fluctuates around 0 in the state of MIX. And the states of S1, S2 and MIX can be classified more easier by being simplified to S1-MIX and S2-MIX two sub-classification cases for the supports of S1 LogBPR and S2 LogBPR are well separated [4]…”

Section: Sac Modulementioning

confidence: 99%

An Adaptive Speech Separation Method Based on Spherical-Regular-Tetrahedral-Microphone Array

Pang¹,

Pei²,

Jiang³

2017

Proceedings of the International Conference on Communication and Electronic Information Engineering (CEIE 2016)

View full text Add to dashboard Cite

For speech separation method, getting the accurate weight vector is very important for the performance of method. Aiming at how to reduce the signal cancellation, this paper focuses on obtaining the weight vector in the speech separation process accurately, and it designs an adaptive speech separation method based on spherical-regular-tetrahedralmicrophone array. The simulation results show that this paper's method can reduce the signal cancellation effectively in speech separation process, so as to have a better separation performance than the similar method base on two-microphone.

show abstract

Section: Sac Modulementioning

confidence: 99%