Design of Automatic Correction System for English Pronunciation Errors based on HMM Algorithm

With the rapid development of English, people's requirements for the level of speech information processing are constantly increasing. However, some errors that cannot be accurately identified often occur in practical applications. In order to effectively improve the translation effect, narrow the scope of translation differences, and reduce the impact of these errors on the results, this article has designed a system that can efficiently and accurately identify the relationship between the content and meaning to be expressed, and can automatically correct these errors. Afterwards, this article inputs the corpus of vowel errors, consonant errors, long and short sound errors, consecutive pronunciation errors and weak pronunciation errors into the correction system for testing, and records the detection and correction results of each pronunciation error. The system can immediately detect and correct pronunciation errors, providing real-time feedback. By cooperating with voice or video input, users can instantly understand their pronunciation problems and make corrections. This immediate feedback not only improves learning effectiveness, but also increases interactivity and autonomy in learning.