Automatic Multilingual Speech Recognition

Nguyen Tuan Anh , Tran Thi Ngoc Linh , Dang Thi Hien

Automatic Speech Recognition (ASR), multi-languages, Vietnamese and Chinese ASR system, LIS-Net model

Automatic Speech Recognition (ASR) for multi-languages is currently attracting more and more attention; however, development is still hampered by the need for language experts. End-to-End ASR simplifies their work by directly predicting the output character based on the acoustic input. This study presents the improvement of LIS-Net model for End-to-End Vietnamese and Chinese ASR system. In this study, an efficient yet accurate end-to-end multilingual multi-speaker ASR model has developed, allowing direct conversion of raw speech audio signals into text of multiple languages. This study proposes a new method of coding labels specifically for multiple languages by pagination labels by language. The results of this study are significantly improved compared to that of baseline models.