Tied Mixture modeling optimization for Korean-digit in the embedded ASR system

Keyphrases

• Triphone 100%
• Speech Recognition System 100%
• Mixture Modeling 100%
• Model Optimization 100%
• Recognition Performance 66%
• Continuous-time Hidden Markov Model 66%
• Representative Experiment 33%
• Performance Degradation 33%
• Modeling Method 33%
• Acoustic Modeling 33%
• Context Model 33%
• Memory Space 33%
• Word Error Rate 33%
• Embedded Systems 33%
• Recognition Rate 33%
• Isolated Word 33%
• Mixture Model 33%
• Semi-continuous 33%
• Task Space 33%
• Remedial Actions 33%
• Speech Task 33%

Computer Science

• Recognition Performance 100%
• Speech Recognition System 100%
• Embedded Systems 50%
• Performance Degradation 50%
• Context Dependent 50%
• Word Error Rate 50%
• Recognition Rate 50%