I am an AI enthusiast and a passionate language learner. While studying at MIT, I specialized in voice recognition, including speech-to-text, text-to-speech, speech-to-emotion, and phoneme recognition. I speak five languages: Chinese, English, French, Japanese, and Spanish. The model I have developed is based on Fairseq technology and fine-tuned on data from 50 different languages. This app primarily focuses on Chinese speech recognition. Through my years of living and working abroad, I’ve observed that many foreigners struggle with accurate pronunciation of Chinese due to the influence of their native languages. Most other educational apps can only tell if the pronunciation is incorrect but fail to point out where the mistake is, making it difficult to correct. This app can recognize tones, initials, and finals in Chinese pronunciation with 98% accuracy, providing fundamental guidance for correct pronunciation. Additionally, the app can identify errors in text expressions and suggest corrections based on native Chinese usage, allowing foreigners to truly learn authentic Chinese

Chinese AI - U Language Coach

Large speech model, supporting recognition of 50 phonemes

Large speech model, supporting recognition of 50 phonemes