Good Neighbors Are All You Need for Chinese Grapheme-to-Phoneme Conversion
This addresses accuracy issues in Chinese text-to-speech and speech recognition systems, representing a domain-specific improvement.
The paper tackles the problem of Chinese grapheme-to-phoneme conversion, where existing systems rely on linguistic knowledge and post-processing that struggle with specific phonetic cases and tone inconsistencies. The proposed Reinforcer method, which emphasizes phonological information between neighboring characters, boosts state-of-the-art architectures by a large margin and demonstrates effectiveness in knowledge transfer scenarios.
Most Chinese Grapheme-to-Phoneme (G2P) systems employ a three-stage framework that first transforms input sequences into character embeddings, obtains linguistic information using language models, and then predicts the phonemes based on global context about the entire input sequence. However, linguistic knowledge alone is often inadequate. Language models frequently encode overly general structures of a sentence and fail to cover specific cases needed to use phonetic knowledge. Also, a handcrafted post-processing system is needed to address the problems relevant to the tone of the characters. However, the system exhibits inconsistency in the segmentation of word boundaries which consequently degrades the performance of the G2P system. To address these issues, we propose the Reinforcer that provides strong inductive bias for language models by emphasizing the phonological information between neighboring characters to help disambiguate pronunciations. Experimental results show that the Reinforcer boosts the cutting-edge architectures by a large margin. We also combine the Reinforcer with a large-scale pre-trained model and demonstrate the validity of using neighboring context in knowledge transfer scenarios.