Ginisupport vector machines for segmental minimum Bayes risk decoding of continuous speech

Keyphrases

• Acoustics 33%
• Approximation Techniques 33%
• Automatic Speech Recognition 33%
• Confidence Score 33%
• Constrained State Estimation 33%
• Continuous Speech 100%
• Continuous Speech Recognition 33%
• Discriminative Model 33%
• Estimation Method 33%
• Feature Space 33%
• Hidden Markov Model 33%
• Hypothesis Combination 33%
• Large Vocabulary Continuous Speech Recognition (LVCSR) 33%
• Limited Vocabulary 33%
• Logistic Regression 66%
• Maximum mutual Information 33%
• Minimum Bayes Risk 100%
• Mutual Information Criterion 33%
• Posterior Distribution 33%
• Problem Recognition 33%
• Recognition Task 33%
• Regression Problem 66%
• Rescoring 33%
• Search Space 33%
• Support Vector Machine 100%
• Vector Machine 100%
• Vocabulary Recognition 33%

Computer Science

• Approximation Technique 33%
• Confidence Score 33%
• Feature Space 33%
• Information Criterion 33%
• Logistic Regression 66%
• Minimum Bayes Risk 100%
• Mutual Information 33%
• Posterior Distribution 33%
• Recognition Problem 33%
• Regression Problem 66%
• Search Space 33%
• Speech Recognition 100%
• Support Vector Machine 100%

Engineering

• Automatic Speech Recognition 33%
• Continuous Speech Recognition 66%
• Feature Space 33%
• Initial Investigation 33%
• Minimum Bayes Risk 100%
• Model System 33%
• Mutual Information 33%
• Posterior Distribution 33%
• Recognition Problem 33%
• Search Space 33%
• Support Vector Machine 100%
• Tasks 33%