This papers investigates two approaches to on-line incremental adaptation of CDHMM parameters. First the popular MAP approach is examined, highlighting difficulties in automatically setting the adaptation rate. To overcome these problems we introduce a new approach based on the multi-observation estimation equations of the forward-backward algorithm called a cumulative likelihood estimate (CLE). Experimental results using these two approaches are compared with and without the use of a speech model for enrolment on isolated word speaker models. In both enrolment procedures, the CLE approach can achieve approximately an EER of 1% for six adaptation sequences using a single digit test token.
Cite as: Yu, K., Mason, J.S. (1996) On-line incremental adaptation for speaker verification using maximum likelihood estimates of CDHMM parameters. Proc. 4th International Conference on Spoken Language Processing (ICSLP 1996), 1752-1755, doi: 10.21437/ICSLP.1996-445
@inproceedings{yu96b_icslp,
author={Kin Yu and John S. Mason},
title={{On-line incremental adaptation for speaker verification using maximum likelihood estimates of CDHMM parameters}},
year=1996,
booktitle={Proc. 4th International Conference on Spoken Language Processing (ICSLP 1996)},
pages={1752--1755},
doi={10.21437/ICSLP.1996-445}
}