R
R. De Mori
Researcher at University of Avignon
Publications - 61
Citations - 2811
R. De Mori is an academic researcher from University of Avignon. The author has contributed to research in topics: Language model & Hidden Markov model. The author has an hindex of 19, co-authored 61 publications receiving 2671 citations. Previous affiliations of R. De Mori include Carnegie Mellon University & McGill University.
Papers
More filters
Proceedings ArticleDOI
Semantic composition process in a speech understanding system
Frédéric Duvert,Marie-Jean Meurs,Christophe Servan,Frédéric Béchet,Fabrice Lefèvre,R. De Mori +5 more
TL;DR: A knowledge representation formalism for SLU is introduced and an automatic interpretation process is described for composing semantic structures from basic semantic constituents using patterns involving constituents and words.
Journal ArticleDOI
Augmenting standard speech recognition features with energy gravity centres
TL;DR: It is shown that the gravity centres of energies in the frequency bands of the first three formants and their first and second time derivatives can be added to the classical set of MFCCs and their second and third time derivatives, resulting in significant performance improvements.
Proceedings ArticleDOI
The use of syllable phonotactics for word hypothesization
R. De Mori,M. Galler +1 more
TL;DR: A search technique incorporating the automatic modeling of lexical variability is introduced for medium or large-vocabulary speaker-independent speech recognition and a new approach for word hypothesization is proposed, based on an acoustic-phonetic unit called the pseudo-syllable segment.
Proceedings Article
Computation of upper-bounds for stochastic context-free languages
TL;DR: This method allows language models to be built based on stochastic context-free grammars and their use with an admissible search algorithm that interprets a speech signal with left-to-right or middle-out strategies.
Proceedings ArticleDOI
Ear-model derived features for automatic speech recognition
TL;DR: The paper provides a theoretical justification that gravity centers in frequency bands computed from zero-crossing information are far more robust to additive telephone noise than GCs computed from FFT spectra.