Class-based speech recognition using a maximum dissimilarity criterion and a tolerance classification margin

Gorin, Arseniy; Jouvet, Denis

doi:10.1109/slt.2012.6424203

Cited by 2 publications

(2 citation statements)

References 11 publications

(15 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Increasing the number of classes decreases the number of available training utterances associated with each class. This problem can be partially handled by soft clustering techniques, such as eigenvoice approach, where the parameters of an unknown speaker are determined as a combination of class models [10], or by explicitly enlarging the class data by allowing one utterance to belong to several classes [9,7].…”

Section: Introductionmentioning

confidence: 99%

Structured GMM Based on Unsupervised Clustering for Recognizing Adult and Child Speech

Gorin

Jouvet

2014

Statistical Language and Speech Processing

Self Cite

View full text Add to dashboard Cite

Speaker variability is a well-known problem of state-of-theart Automatic Speech Recognition (ASR) systems. In particular, handling children speech is challenging because of substantial differences in pronunciation of the speech units between adult and child speakers. To build accurate ASR systems for all types of speakers Hidden Markov Models with Gaussian Mixture Densities were intensively used in combination with model adaptation techniques. This paper compares different ways to improve the recognition of children speech and describes a novel approach relying on Class-Structured Gaussian Mixture Model (GMM). A common solution for reducing the speaker variability relies on gender and age adaptation. First, it is proposed to replace gender and age by unsupervised clustering. Speaker classes are first used for adaptation of the conventional HMM. Second, speaker classes are used for initializing structured GMM, where the components of Gaussian densities are structured with respect to the speaker classes. In a first approach mixture weights of the structured GMM are set dependent on the speaker class. In a second approach the mixture weights are replaced by explicit dependencies between Gaussian components of mixture densities (as in stranded GMMs, but here the GMMs are class-structured). The different approaches are evaluated and compared on the TIDIG-ITS task. The best improvement is achieved when structured GMM is combined with feature adaptation.

show abstract

Section: Introductionmentioning

confidence: 99%

Structured GMM Based on Unsupervised Clustering for Recognizing Adult and Child Speech

Gorin

Jouvet

2014

Statistical Language and Speech Processing

Self Cite

View full text Add to dashboard Cite

show abstract

“…With respect to the training process, increasing the number of classes decreases the number of utterances associated with each class. This problem can be partially handled by soft clustering techniques, such as eigenvoice approach, where the parameters of an unknown speaker are determined as a combination of class models [7], or by explicitly enlarging the class-associated data by allowing one utterance to belong to several classes [8,9].…”

Section: Introductionmentioning

confidence: 99%

Component structuring and trajectory modeling for speech recognition

Gorin¹,

Jouvet²

2014

Interspeech 2014

Self Cite

View full text Add to dashboard Cite

When the speech data are produced by speakers of different age and gender, the acoustic variability of any given phonetic unit becomes large, which degrades speech recognition performance. A way to go beyond the conventional Hidden Markov Model is to explicitly include speaker class information in the modeling. Speaker classes can be obtained by unsupervised clustering of the speech utterances.This paper introduces a structuring of the Gaussian components of the GMM densities with respect to speaker classes. In a first approach, the structuring of the Gaussian components is combined with speaker class-dependent mixture weights. In a second approach, the structuring is used with mixture transition matrices, which add dependencies between Gaussian components of mixture densities (as in stranded GMMs). The different approaches are evaluated and compared in detail on the TIDIG-ITS task. Significant improvements are obtained using the proposed approaches based on structured components. Additional results are reported for phonetic decoding on the NEOLOGOS database, a large corpus of French telephone data.

show abstract

Class-based speech recognition using a maximum dissimilarity criterion and a tolerance classification margin

Cited by 2 publications

References 11 publications

Structured GMM Based on Unsupervised Clustering for Recognizing Adult and Child Speech

Structured GMM Based on Unsupervised Clustering for Recognizing Adult and Child Speech

Component structuring and trajectory modeling for speech recognition

Contact Info

Product

Resources

About