We describe the use of Support Vector Machines for phonetic classi- fication on the TIMIT corpus. Unlike previous work, in which entire phonemes are classified, our system operates in a framewise manner and is intended for use as the front-end of a hybrid system similar to ABBOT. We therefore avoid the problems of classifying variablelength vectors. Our frame-level phone classification accuracy on the complete TIMIT test set is competitive with other results from the literature. In addition, we address the serious problem of scaling Support Vector Machines by using the Kernel Fisher Discriminant.
Cite as: Salomon, J., King, S., Salomon, J. (2002) Framewise phone classification using support vector machines. Proc. 7th International Conference on Spoken Language Processing (ICSLP 2002), 2645-2648, doi: 10.21437/ICSLP.2002-352
@inproceedings{salomon02_icslp,
author={Jesper Salomon and Simon King and Jesper Salomon},
title={{Framewise phone classification using support vector machines}},
year=2002,
booktitle={Proc. 7th International Conference on Spoken Language Processing (ICSLP 2002)},
pages={2645--2648},
doi={10.21437/ICSLP.2002-352},
issn={2958-1796}
}