Publications de Subhadeep Dey | EPFL Graph Search

Incremental Semi-Supervised Learning for Multi-Genre Speech Recognition

Hervé Bourlard, Petr Motlicek, Subhadeep Dey

2020

A Bayesian Approach To Inter-Task Fusion For Speaker Recognition

In i-vector based speaker recognition systems, back-end classifiers are trained to factor out nuisance information and retain only the speaker identity. As a result, variabilities arising due to gender, language and accent ( among many others) are suppress ...

IEEE2019

End-to-end text-dependent speaker verification using novel distance measures

Petr Motlicek, Subhadeep Dey

This paper explores novel ideas in building end-to-end deep neural network (DNN) based text-dependent speaker verification (SV) system. The baseline approach consists of mapping a variable length speech segment to a fixed dimensional speaker vector by esti ...

2018

Analysis of Language Dependent Front-End for Speaker Recognition

Petr Motlicek, Subhadeep Dey

In Deep Neural Network (DNN) i-vector based speaker recognition systems, acoustic models trained for Automatic Speech Recognition are employed to estimate sufficient statistics for i-vector modeling. The DNN based acoustic model is typically trained on a w ...

ISCA-INT SPEECH COMMUNICATION ASSOC2018

Phonetic aware techniques for Speaker Verification

Subhadeep Dey

The goal of this thesis is to improve current state-of-the-art techniques in speaker verification (SV), typically based on â identity-vectorsâ (i-vectors) and deep neural network (DNN), by exploiting diverse (phonetic) information extracted using variou ...

EPFL2018

Template-matching for text-dependent speaker verification

Petr Motlicek, Subhadeep Dey

In the last decade, i-vector and Joint Factor Analysis (JFA) approaches to speaker modeling have become ubiquitous in the area of automatic speaker recognition. Both of these techniques involve the computation of posterior probabilities, using either Gauss ...

2017

Exploiting sequence information for text-dependent Speaker Verification

Petr Motlicek, Subhadeep Dey

Model-based approaches to Speaker Verification (SV), such as Joint Factor Analysis (JFA), i-vector and relevance Maximum-a-Posteriori (MAP), have shown to provide state-of-the-art performance for text-dependent systems with fixed phrases. The performance o ...

Ieee2017

INTRA-CLASS COVARIANCE ADAPTATION IN PLDA BACK-ENDS FOR SPEAKER VERIFICATION

Petr Motlicek, Subhadeep Dey

Multi-session training conditions are becoming increasingly common in recent benchmark datasets for both text-independent and text-dependent speaker verification. In the state-of-the-art i-vector framework for speaker verification, such conditions are addr ...

Ieee2017

Information theoretic clustering for unsupervised domain-adaptation

Petr Motlicek, Subhadeep Dey

The aim of the domain-adaptation task for speaker verification is to exploit unlabelled target domain data by using the labelled source domain data effectively. The i-vector based Probabilistic Linear Dis- criminant Analysis (PLDA) framework approaches thi ...

IEEE2016

Deep neural network based posteriors for text-dependent speaker verification

Petr Motlicek, Subhadeep Dey

The i-vector and Joint Factor Analysis (JFA) systems for text- dependent speaker verification use sufficient statistics computed from a speech utterance to estimate speaker models. These statis- tics average the acoustic information over the utterance ther ...

IEEE2016