Learn More
This paper accounts for the DEMOSTHENES application, an interactive tool for the correction of foreign language learners' pronunciation. After giving the didactic and technical arguments advocating for DEMOSTHENES, the authors describe the phase of creation and labelling of a sound database for the application language (Dutch). The methodolgy for the(More)
There are two major approaches to speech recognition: frame-based and segment-based approach. Frame-based approach, e.g. HMM, assumes statistical independence and identical distribution of observation in each state. In addition it incorporates weak duration constrains. Segmentbased approach is computational expensive and rough modelling easilly occurs if(More)
The aim of this work is to provide a proper framework for the simulation and the optimization of the event building, the on-line third level trigger, and complete event reconstruction processor farm for the future HERA-B experiment. A discrete event, process oriented, simulation developed in concurrent μC++ is used for modelling the farm nodes running with(More)
| In this paper we report a series of tests carried out on our hybrid HMM/ANN systems which aims at combining Neural Networks theory and Hidden Markov Models (HMMs) for speech recognition of a continuous speech French database: BREF-80. As this database is not manually labelled , we describe a new method based on the temporal alignment of the speech signal(More)
The concept of mixtures of discrete HMMs (MDHMM) is introduced. The application of MDHMMs to the classification of mixtures of signals is described. The optimal decision rule is presented. Alternative algorithms with reduced computational load are proposed: a simplified decision statistic is defined and sub-optimal search methods are discussed. The(More)