Evaluating Logit-Based GOP Scores for Mispronunciation Detection
Evaluating Logit-Based GOP Scores for Mispronunciation Detection
Pronunciation assessment relies on goodness of pronunciation (GOP) scores, traditionally derived from softmax-based posterior probabilities. However, posterior probabilities may suffer from overconfidence and poor phoneme separation, limiting their effectiveness. This study compares logit-based GOP scores with probability-based GOP scores for mispronunciation detection. We conducted our experiment on two L2 English speech datasets spoken by Dutch and Mandarin speakers, assessing classification performance and correlation with human ratings. Logit-based methods outperform probability-based GOP in classification, but their effectiveness depends on dataset characteristics. The maximum logit GOP shows the strongest alignment with human perception, while a combination of different GOP scores balances probability and logit features. The findings suggest that hybrid GOP methods incorporating uncertainty modeling and phoneme-specific weighting improve pronunciation assessment.
Aditya Kamlesh Parikh、Cristian Tejedor-Garcia、Catia Cucchiarini、Helmer Strik
语言学常用外国语汉语
Aditya Kamlesh Parikh,Cristian Tejedor-Garcia,Catia Cucchiarini,Helmer Strik.Evaluating Logit-Based GOP Scores for Mispronunciation Detection[EB/OL].(2025-07-08)[2025-07-16].https://arxiv.org/abs/2506.12067.点此复制
评论