Laurent Couvreur

TL;DR: It is shown that chirp group delay representations are potentially useful for improving ASR performance and presented one application in feature extraction for automatic speech recognition (ASR), which can be guaranteed to be spike-free.

TL;DR: The proposed method relies on data coming from a contour segmentation technique, which extracts an expression skeleton of facial features and derives simple distance coefficients from every face image of a video sequence and demonstrates the feasibility of facial expression classification with simple data.

TL;DR: The proposed model selection approach is shown to improve significantly recognition accuracy for a connected digit task in both simulated and real reverberant environments, outperforming standard channel normalization techniques.

TL;DR: It is shown that two of the representations proposed perform better, contain equivalent or complementary information to that of the power spectrum and are potentially useful for improving ASR performance.

TL;DR: This paper presents a system for classifying facial expressions based on a data fusion process relying on the Belief Theory, and shows that the BeT classifier outperforms both the BaT and HMM classifiers for the considered application.