The BKA voice comparison system SPES is designed for forensic examination of speech recordings. The classical GMM-UBM framework based on MAP adaptation as described by Reynolds et al. is extended by the generation of recording adapted background models (RABMs). We present results from experiments using real case data. These results show how the most critical properties of real case recordings such as duration, channel, and samples per speaker influence system performance.
Cite as: Becker, T., Jessen, M., Alsbach, S., Broß, F., Meier, T. (2010) SPES: The BKA Forensic Automatic Voice Comparison System. Proc. The Speaker and Language Recognition Workshop (Odyssey 2010), paper 11
@inproceedings{becker10_odyssey,
author={Timo Becker and Michael Jessen and Sebastian Alsbach and Franz Broß and Torsten Meier},
title={{SPES: The BKA Forensic Automatic Voice Comparison System}},
year=2010,
booktitle={Proc. The Speaker and Language Recognition Workshop (Odyssey 2010)},
pages={paper 11}
}