English Speech Recognition and Pronunciation Quality Evaluation Model Based on Neural Network
Table 1
Manual evaluation grades and evaluation criteria.
Grade
Intonation
Speed of speech
Rhythm
Intonation
Overall
A
The content is complete and accurate, the voice is clear, the whole body is smooth, and there are no obvious pronunciation errors.
Moderate speech rate
Accurate accent pronunciation and strong sense of rhythm
Accurate and natural intonation
Excellent pronunciation in general
B
The content is relatively complete and accurate, the pronunciation is relatively clear and fluent, and there are no serious pronunciation errors.
Speak faster (slow)
Accent pronunciation is more accurate.
The tone is more accurate and natural
Pronunciation is generally good
C
The text is nearly complete with almost clear and fluent pronunciation, and contains word errors that interfere with comprehension.
Speaking too fast (slow)
Good sense of rhythm
The intonation is basically accurate, but not natural enough
General grasp of pronunciation
D
The content is incomplete and accurate, the pronunciation is not clear and fluent, and there are serious pronunciation errors that affect understanding.
Speaking too fast (slow)
Mispronunciation of accents, too many (less) accents, and poor sense of rhythm
