Spectro-Temporal Gabor Features for Speaker Recognition

TitleSpectro-Temporal Gabor Features for Speaker Recognition
Publication TypeConference Paper
Year of Publication2012
AuthorsLei, H., Meyer B. T., & Mirghafori N.
Page(s)4241-4244
Other Numbers3239
Abstract

In this work, we have investigated the performance of 2D Gaborfeatures (known as spectro-temporal features) for speaker recognition.Gabor features have been used mainly for automatic speechrecognition (ASR), where they have yielded improvements. We exploreddifferent Gabor feature implementations, along with differentspeaker recognition approaches, on ROSSI [1] and NIST SRE08databases. Using the noisy ROSSI database, the Gabor features performedas well as the MFCC features standalone, and score-level

Acknowledgment

This work is sponsored by Air Force Research Laboratory undercontract FA8750-10-C-0214.

URLhttp://www.icsi.berkeley.edu/pubs/speech/gaborfeatures12.pdf
Bibliographic Notes

Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2012), pp. 4241-4244, Kyoto, Japan

Abbreviated Authors

H. Lei, B. T. Meyer, and N. Mirghafori

ICSI Research Group

Speech

ICSI Publication Type

Article in conference proceedings