Mathieu Ben

A Monte-Carlo method for score normalization in Automatic Speaker Verification using Kullback-Leibler distances

In this paper, we propose a new score normalization technique in Automatic Speaker Verification (ASV): the D-Norm. The main advantage of this score normalization is that it does not need any additional speech data nor external speaker population, as opposed to the state-of-the-art approaches. The D-Norm is based on the use of Kullback-Leibler (KL) distances in an ASV context. In a first step. we estimate the KL distances with a Monte-Carlo method and we experimentally show that they are correlated with the verification scores. In a second step, we use this correlation to implement a score normalization procedure, the D-Norm. We analyse its performance and we compare it to that of a conventional normalization, the Z-Norm. The results show that performance of the D-Norm is comparable to that of the Z-Norm. We then conclude about the results we obtain and we discuss the applications of this work.

D-MAP: a distance-normalized MAP estimation of speaker models for automatic speaker verification

We introduce a MAP estimation of speaker models in automatic speaker verification with a distance constraint: the D-MAP adaptation. The D-MAP is based on the Kullback-Leibler distances and provides an easy way to automatically compute a speaker-dependent adaptation of the model parameters. We formulate a distance constrained MAP criterion and we show an equivalence between the D-MAP adaptation and the score normalization called D-norm. From the results obtained with the D-MAP technique, we show that this method gives better performance than a classical speaker-independent MAP adaptation. It is also found that the D-MAP based system without score normalization performs similarly to a classical MAP system with a model-based score normalization.

Improving cluster selection and event modeling in unsupervised mining for automatic audiovisual video structuring

Can we discover audio-visually consistent events from videos in a totally unsupervised manner? And, how to mine videos with different genres? In this paper we present our new results in automatically discovering audio-visual events. A new measure is proposed to select audio-visually consistent elements from the two dendrograms respectively representing hierarchical clustering results for the audio and visual modalities. Each selected element corresponds to a candidate event. In order to construct a model for each event, each candidate event is represented as a group of clusters, and a voting mechanism is applied to select training examples for discriminative classifiers. Finally, the trained model is tested on the entire video to select video segments that belong to the event discovered. Experimental results on different and challenging genres of videos, show the effectiveness of our approach.