This paper demonstrates the robustness of group-delay based features for speech processing. An analysis of group delay functions is presented which show that these features retain formant structure even in noise. Furthermore, a speaker verification task performed on the NIST 2003 database show lesser error rates, when compared with the traditional MFCC features. We also mention about using feature diversity to dynamically choose the feature for every claimed speaker.
Cite as: Padmanabhan, R., Parthasarathi, S.H.K., Murthy, H.A. (2009) Robustness of phase based features for speaker recognition. Proc. Interspeech 2009, 2355-2358, doi: 10.21437/Interspeech.2009-397
@inproceedings{padmanabhan09_interspeech,
author={R. Padmanabhan and Sree Hari Krishnan Parthasarathi and Hema A. Murthy},
title={{Robustness of phase based features for speaker recognition}},
year=2009,
booktitle={Proc. Interspeech 2009},
pages={2355--2358},
doi={10.21437/Interspeech.2009-397}
}