论文信息 - Applying Restrained Likelihood and Floating TMR to Multi-Speaker Identification for Co-channel Speech

Applying Restrained Likelihood and Floating TMR to Multi-Speaker Identification for Co-channel Speech

In this paper, a multi-speaker identification system for co-channel speech is proposed. By using constrained likelihood and floating TMR method, this system can identify two speakers on co-channel speech with high accuracy.

Peng Li | Bo Xu | Wenju Liu | Yong Guan | Xueliang Zhang

[1] DeLiang Wang,et al. Robust Speaker Recognition Using Binary Time-Frequency Masks , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.

[2] Sam T. Roweis,et al. Factorial models and refiltering for speech separation and denoising , 2003, INTERSPEECH.

[3] Sam T. Roweis,et al. One Microphone Source Separation , 2000, NIPS.

[4] John R. Hershey,et al. Super-human multi-talker speech recognition: the IBM 2006 speech separation challenge system , 2006, INTERSPEECH.

[5] DeLiang Wang,et al. Co-channel speaker identification using usable speech extraction based on multi-pitch tracking , 2003, 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03)..