LSTM Acoustic Models Learn to Align and Pronounce with Graphemes
arXiv:2008.06121
Abstract
Automated speech recognition coverage of the world's languages continues to expand. However, standard phoneme based systems require handcrafted lexicons that are difficult and expensive to obtain. To address this problem, we propose a training methodology for a grapheme-based speech recognizer that can be trained in a purely data-driven fashion. Built with LSTM networks and trained with the cross-entropy loss, the grapheme-output acoustic models we study are also extremely practical for real-world applications as they can be decoded with conventional ASR stack components such as language models and FST decoders, and produce good quality audio-to-grapheme alignments that are useful in many speech applications. We show that the grapheme models are competitive in WER with their phoneme-output counterparts when trained on large datasets, with the advantage that grapheme models do not require explicit linguistic knowledge as an input. We further compare the alignments generated by the phoneme and grapheme models to demonstrate the quality of the pronunciations learnt by them using four Indian languages that vary linguistically in spoken and written forms.
arXiv comment5 pages, 4 figures. This work was done between summer 2018 and spring 2019
References in corpus (5)
- Attention-Based Models for Speech Recognition
- Tacotron: Towards End-to-End Speech Synthesis
- End-to-End ASR-free Keyword Search from Speech
- Neural Speech Recognizer: Acoustic-to-Word LSTM Model for Large Vocabulary Speech Recognition
- Direct Acoustics-to-Word Models for English Conversational Speech Recognition