Jean-François Bonastre

Transfer Function-Based Voice Transformation for Speaker Recognition

This paper investigates the effect of a transfer function-based voice transformation on automatic speaker recognition system performance. We focus on increasing the impostor acceptance rate, by modifying the voice of an impostor in order to target a specific speaker. This paper is based on the following idea: in several applications and particularly in forensic situations, it is reasonable to think that some organizations have a knowledge on the speaker recognition method used and could impersonate a given, well known speaker. We also evaluate the effect of the voice transformation when the transformation is applied both on client and impostor trials. This paper presents some experiments based on NIST SRE 2005 protocol. The results show that the voice transformation allows a drastic increase of the false acceptance rate, without damaging the natural aspect of the voice. It seems also that this kind of voice transformation could be efficient for reducing the inter-session mismatch

Modeling nuisance variabilities with factor analysis for GMM-based audio pattern classification

Abstract: Audio pattern classification represents a particular statistical classification task and includes, for example, speaker recognition, language recognition, emotion recognition, speech recognition and, recently, video genre classification. The feature being used in all these tasks is generally based on a short-term cepstral representation. The cepstral vectors contain at the same time useful information and nuisance variability, which are difficult to separate in this domain. Recently, in the context of GMM-based recognizers, a novel approach using a Factor Analysis (FA) paradigm has been proposed for decomposing the target model into a useful information component and a session variability component. This approach is called Joint Factor Analysis (JFA), since it models jointly the nuisance variability and the useful information, using the FA statistical method. The JFA approach has even been combined with Support Vector Machines, known for their discriminative power. In this article, we successfully apply this paradigm to three automatic audio processing applications: speaker verification, language recognition and video genre classification. This is done by applying the same process and using the same free software toolkit. We will show that this approach allows for a relative error reduction of over 50% in all the aforementioned audio processing tasks.

Accurate Log-Likelihood Ratio Estimation by using Test Statistical Model for Speaker Verification

In this paper we propose an accurate estimation of the log-likelihood ratio (LLR) thanks to a statistical modelling of the test data. This work takes place within the framework of GMM/UBM based speaker verification. Modelling the test data using a statistical model like a GMM shows several advantages, and particularly it allows to reduce the influence of out-of-domain data thanks to the underlined statistical model. In this paper, we explore the interests of such methods, using a GMM modelling of the test data. We propose also an extension of this approach to the MAP-based speaker model adaptation. Some experiments based on the NIST SRE 2005 protocol are presented and show a significant gain (between 4% and 5% in relative compared to our NIST GMM/UBM baseline) by using our LLR estimation

Constrained discriminative speaker verification specific to normalized i-vectors.

This paper focuses on discriminative trainings (DT) applied to i-vectors after Gaussian probabilistic linear discriminant analysis (PLDA). If DT has been successfully used with non-normalized vectors, this technique struggles to improve speaker detection when i-vectors have been first normalized, whereas the latter option has proven to achieve best performance in speaker verification. We propose an additional normalization procedure which limits the amount of coefficient to discriminatively train, with a minimal loss of accuracy. Adaptations of logistic regression based-DT to this new configuration are proposed, then we introduce a discriminative classifier for speaker verification which is a novelty in the field.

Local binary patterns as features for speaker recognition.

The i-vector framework witnessed great success in the past years in speaker recognition (SR). The feature extraction process is central in SR systems and many features have been developed over the years to improve the recognition performance. In this paper, we present a new feature representation which borrows a concept initially developed in computer vision to characterize textures called Local Binary Patterns (LBP). We explore the use of LBP as features for speaker recognition and show that using them as descriptors for cepstral coefficients dynamics (replacing ∆ and ∆∆ in the regular MFCC representation) results in more efficient features and yield up to 15% of relative improvement compared to the baseline system performance in both clean and noisy conditions. keywords: local binary patterns, feature extraction, ivector.

Iterative Bayesian and MMSE-based noise compensation techniques for speaker recognition in the i-vector space.

Dealing with additive noise in the i-vector space can be challenging due to the complexity of its effect in that space. Several compensation techniques have been proposed in the last years to either remove the noise effect by setting a noise model in the i-vector space or build better scoring techniques that take environment perturbations into account. We recently presented a new efficient Bayesian cleaning technique operating in the ivector domain named I-MAP that improves the baseline system performance by up to 60%. This technique is based on Gaussian models for the clean and noise i-vectors distributions. After IMAP transformation, these hypothesis are probably less correct. For this reason, we propose to apply another MMSE-based approach that uses the Kabsch algorithm. For a certain noise, it estimates the best translation vector and rotation matrix between a set of train noisy i-vectors and their clean counterparts based on RMSD criterion. This transformation is then applied on noisy test i-vectors in order to remove the noise effect. We show that applying the Kabsch algorithm allows to reach a 40% relative improvement in EER(%) compared to a baseline system performance and that, when combined with I-MAP and repeated iteratively, it allows to reach 85% of relative improvement. keywords: i-vector, additive noise, Kabsch algorithm, IMAP

Impact of rhythm on forensic voice comparison reliability

It is common to see voice recordings being presented as a forensic trace in court. Generally, a forensic expert is asked to analyze both suspect and criminals voice samples in order to indicate whether the evidence supports the prosecution (same-speaker) or defence (different-speakers) hypotheses. This process is known as Forensic Voice Comparison (FVC). Since the emergence of the DNA typing model, the likelihood-ratio (LR) framework has become the new golden standard in forensic sciences. The LR not only supports one of the hypotheses but also quantifies the strength of its support. However, the LR accepts some practical limitations due to its estimation process itself. It is particularly true when Automatic Speaker Recognition (ASpR) systems are considered as they are outputting a score in all situations regardless of the case specific conditions. Indeed, several factors are not taken into account by the estimation process like the quality and quantity of information in both voice recordings, their phonological content or also the speakers intrinsic characteristics, etc. All these factors put into question the validity and reliability of FVC. In our recent study, we showed that intra-speaker variability explains 2/3 of the system losses. In this article, we investigate the relations between intra-speaker variability and rhythmic parameters.