g2pM: A Neural Grapheme-to-Phoneme Conversion Package for Mandarin Chinese Based on a New Open Benchmark Dataset

Conversion of Chinese graphemes to phonemes (G2P) is an essential component\nin Mandarin Chinese Text-To-Speech (TTS) systems. One of the biggest challenges\nin Chinese G2P conversion is how to disambiguate the pronunciation of\npolyphones - characters having multiple pronunciations. Although many academic\nefforts have been made to address it, there has been no open dataset that can\nserve as a standard benchmark for fair comparison to date. In addition, most of\nthe reported systems are hard to employ for researchers or practitioners who\nwant to convert Chinese text into pinyin at their convenience. Motivated by\nthese, in this work, we introduce a new benchmark dataset that consists of\n99,000+ sentences for Chinese polyphone disambiguation. We train a simple\nneural network model on it, and find that it outperforms other preexisting G2P\nsystems. Finally, we package our project and share it on PyPi.\n