Maximum likelihood Linear Dimension Reduction of heteroscedastic feature for robust Speaker Recognition

Keyphrases

• Covariance 25%
• Dimensionality Reduction 50%
• Fusion Approach 25%
• Heteroscedastic 100%
• Heteroscedasticity 25%
• Homoscedasticity 75%
• I-vector 25%
• Indian Institute of Technology 25%
• Linear Discriminant Analysis 75%
• Maximum Likelihood 100%
• National Institute of Standards 25%
• Nonlinear Dimensionality Reduction 100%
• Robust Forensics 25%
• Robust Speaker Recognition 100%
• Speaker Recognition Evaluation 25%
• Speaker Recognition System 25%
• Speech Utterances 25%
• Telephone 25%

Computer Science

• Linear Discriminant Analysis 100%
• maximum-likelihood 100%
• Principal Components 33%
• Recognition System 33%
• Speaker Recognition 100%
• Telephone 33%

Mathematics

• Covariance 33%
• Discriminant Analysis 100%
• Heteroscedasticity 33%
• Maximum Likelihood 100%
• Principal Components 33%