Speech recognition in adverse environment has attracted the attention of many researchers. Unfortunately performances seem to be too poor, highly dependent on the nature of noise and, of course, on the signal to noise ratio. In this paper we describe the design and the realisation of an assessment experiment on the effect of additive noise, in the simple task of speaker dependent digits recognition for Italian language. A total of four recognisers have been tested. Four speakers, three kind of noises and six different signal to noise ratios have been used. The assessment has been carried on according to the directions, procedures and reference databases produced in the ESPRIT 2589 (SAM) project.
Cite as: Falcone, M., Daino, S., Ragazzini, S., Ceglie, F. (1992) Assessing recognisers for Italian digits recognition for added noise signals. Proc. ETRW on Speech Processing in Adverse Conditions, 219-222
@inproceedings{falcone92_spac,
author={M. Falcone and S. Daino and S. Ragazzini and F. Ceglie},
title={{Assessing recognisers for Italian digits recognition for added noise signals}},
year=1992,
booktitle={Proc. ETRW on Speech Processing in Adverse Conditions},
pages={219--222}
}