Add to ORKGSpeaker recognition using support vector machines (SVMs) with features derived from generative models has been shown to perform well. Typically, a universal background model (UBM) is adapted to each utterance yielding a set of features that are used in an SVM. We consider the case where the UBM is a Gaussian mixture model (GMM), and maximum likelihood linear regression (MLLR) adap-tation is used to adapt the means of the UBM. Recent work has examined this setup for the case where a global MLLR transform is applied to all the mixture components of the GMM UBM. This work produced positive results that warrant examining this setup with multi-class MLLR adaptation, which groups the UBM mixture components into classes and applies a different tra...
Abstract. Generative Gaussian Mixture Models (GMMs) are known to be the dominant approach for modeli...
This paper extends the within-class covariance normalization (WCCN) technique described in [1, 2] fo...
International audienceMost state-of-the-art speaker recognition systems are partially or completely ...
Speaker recognition using support vector machines (SVMs) with features derived from generative model...
Abstract—We present a new modeling approach for speaker recognition that uses the maximum-likelihood...
This paper describes a speaker recognition system based on feature extraction utilizing the constrai...
The goal of this thesis is to find new and efficient features for speaker recognition. We are mostly...
Most of state-of-the-art speaker recognition systems are based on Gaussian Mixture Models (GMM), tra...
In this paper, we present a newmodeling approach for speaker recognition, which uses a kind of novel...
International audienceIn this paper, we extend the recently introduced Maximum Like- lihood Linear R...
Gaussian mixture model (GMM) supervector is one of the effective techniques in text independent spea...
ii Speaker adaptation based on the Universal Background Model (UBM) has become a standard approach f...
We use a multi-layer perceptron (MLP) to transform cep-stral features into features better suited fo...
In recent years, adaptation techniques have been given special focus in speaker recognition tasks, m...
Recent research has demonstrated the merit of combining Gaussian mixture models and support vector m...
Abstract. Generative Gaussian Mixture Models (GMMs) are known to be the dominant approach for modeli...
This paper extends the within-class covariance normalization (WCCN) technique described in [1, 2] fo...
International audienceMost state-of-the-art speaker recognition systems are partially or completely ...
Speaker recognition using support vector machines (SVMs) with features derived from generative model...
Abstract—We present a new modeling approach for speaker recognition that uses the maximum-likelihood...
This paper describes a speaker recognition system based on feature extraction utilizing the constrai...
The goal of this thesis is to find new and efficient features for speaker recognition. We are mostly...
Most of state-of-the-art speaker recognition systems are based on Gaussian Mixture Models (GMM), tra...
In this paper, we present a newmodeling approach for speaker recognition, which uses a kind of novel...
International audienceIn this paper, we extend the recently introduced Maximum Like- lihood Linear R...
Gaussian mixture model (GMM) supervector is one of the effective techniques in text independent spea...
ii Speaker adaptation based on the Universal Background Model (UBM) has become a standard approach f...
We use a multi-layer perceptron (MLP) to transform cep-stral features into features better suited fo...
In recent years, adaptation techniques have been given special focus in speaker recognition tasks, m...
Recent research has demonstrated the merit of combining Gaussian mixture models and support vector m...
Abstract. Generative Gaussian Mixture Models (GMMs) are known to be the dominant approach for modeli...
This paper extends the within-class covariance normalization (WCCN) technique described in [1, 2] fo...
International audienceMost state-of-the-art speaker recognition systems are partially or completely ...