Pronunciation-aware unique character encoding for RNN Transducer-based Mandarin speech recognition
This work addresses a domain-specific issue in Mandarin ASR by offering an incremental improvement to handle homophones more effectively.
The authors tackled the homophone problem in Mandarin speech recognition by proposing a pronunciation-aware unique character encoding for RNN Transducer models, which improved recognition accuracy on Aishell and MagicData datasets.
For Mandarin end-to-end (E2E) automatic speech recognition (ASR) tasks, compared to character-based modeling units, pronunciation-based modeling units could improve the sharing of modeling units in model training but meet homophone problems. In this study, we propose to use a novel pronunciation-aware unique character encoding for building E2E RNN-T-based Mandarin ASR systems. The proposed encoding is a combination of pronunciation-base syllable and character index (CI). By introducing the CI, the RNN-T model can overcome the homophone problem while utilizing the pronunciation information for extracting modeling units. With the proposed encoding, the model outputs can be converted into the final recognition result through a one-to-one mapping. We conducted experiments on Aishell and MagicData datasets, and the experimental results showed the effectiveness of the proposed method.