Automatic speaker’s role classification with a bottom-up acoustic feature selection

Keyphrases

• Acoustic Features 100%
• Speaker Role Recognition 100%
• Acoustic Feature Selection 100%
• Classification Accuracy 50%
• Task Set 50%
• Map Task 50%
• Speech Signal 50%
• Classification Results 50%
• Interactive Dialogue 50%
• Gender-based Classification 50%
• Statistical Features 50%
• Dimension Features 50%

Computer Science

• Acoustic Feature 100%
• Feature Selection 100%
• Feature Extraction 100%
• classification result 33%
• Feature Dimension 33%