Denis Jouvet

TL;DR: In this paper, the speech signal is modelled by means of a hidden Markov model and, at each instant t: equalization filters are constituted in association with the paths in the Markov sense at instant t; at least a plurality of the equalization filtering filters are applied to the sound frames to obtain, at instant T, an utterance probability for each of the paths respectively associated with the equalisation filters applied, and the filtered frame supplied by the selected equalization filter is selected as the equalized frame.

TL;DR: A noise-robust front- end designed within a collaboration of Motorola, France Télécom and Alcatel for the ETSI standardization of the advanced front-end for distributed speech recognition (DSR).

TL;DR: The results show that the training process is quite robust to some errors in the pronunciation lexicon, whereas pronunciation Lexicon errors are harmful in the decoding process.

TL;DR: The results obtained prove that HMM adaptation and preprocessing techniques can be advantageously combined to improve Automatic Speech Recognition (ASR) robustness and show that spectral subtraction improves speech detection under noisy GSM conditions.

TL;DR: This work proposes an approach to grapheme-to-phoneme conversion based on a probabilistic method: Conditional Random Fields (CRF), which compares favorably with the performance of the state-of-the-art Joint-Multigram Models for the quality of the pronunciations, but provides better recall and precision measures for multiple pronunciation variants generation.