[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Features for robust speaker identification

One feature we proposed and found to be rather effective for robust speaker identification is GFCC (gammatone frequency cepstral coefficient). Its description and analysis are given below:
- Shao Y. and Wang D.L. (2008): "Robust speaker identification using auditory features and computational auditory scene analysis." ICASSP-08, pp. 1589-1592.
- Zhao X., Shao Y., and Wang D.L. (2012): "CASA-based robust speaker identification," IEEE Transactions on Audio, Speech, and Language Processing, vol. 20, pp. 1608-1616.
- Zhao X. and Wang D.L. (2013): "Analyzing noise robustness of MFCC and GFCC features in speaker identification," ICASSP-13, pp. 7204-7208. 

You can also find the Matlab code for GFCC extraction on my lab's website.

Cheers,
DeLiang

On 9/16/2014 12:23 PM, Celestino Alvarez wrote:

Dear list,
I was planning to build a speaker identification application, and I was wondering what are the best features for a robust identification. 

Any advise on the right papers to read, would help.

Best,

Tino


--
------------------------------------------------------------
DeLiang Wang, Professor
Co-Editor-in-Chief, Neural Networks
Department of Computer Science and Engineering
The Ohio State University
2015 Neil Ave.
Columbus, OH 43210-1277, U.S.A.

Phone: 614-292-6827 (OFFICE); 614-292-7402 (LAB)
http://www.cse.ohio-state.edu/~dwang

"Happiness = Reality - Expectation"