Assessment of Non-Native Speech Intelligibility using Wav2vec2-based Mispronunciation Detection and Multi-level Goodness of Pronunciation Transformer

Keyphrases

• Speech Intelligibility 100%
• Non-native Speech 100%
• Transformer 100%
• Pronunciation Assessment 100%
• Mispronunciation Detection 100%
• Wav2vec 2.0 100%
• Goodness of Pronunciation 100%
• Mispronunciation Diagnosis 66%
• Phonetic Level 66%
• Detection System 33%
• Non-native 33%
• Second Language 33%
• Suprasegmental Features 33%
• Language Learners 33%
• Feature-based 33%
• Listener Perception 33%
• Performance-based 33%
• L2 Learners 33%
• Giving Feedback 33%
• Self-directed 33%
• Pronunciation Feature 33%
• Challenging Problems 33%
• L2 Speech 33%
• Diagnosis System 33%
• Phoneme Recognition 33%
• L2 English Speech 33%
• Multi-granular 33%
• Automatic Pronunciation Assessment 33%
• Speech Data 33%
• Speech Dataset 33%
• Feature Perception 33%
• Feature Granularity 33%
• Segmental Phonemes 33%
• Computer-assisted Pronunciation Training 33%
• Pronunciation Quality 33%

Computer Science

• Language Learner 100%
• Granularity 100%
• Assessment Tool 100%
• Computer-Assisted Pronunciation Training 100%

Physics

• Phoneme 100%