Most end-to-end speech recognition systems model text directly as a sequence of characters or sub-words. Current approaches to sub-word extraction only consider character sequence frequencies, which at times produce inferior sub-word segmentation that might lead to erroneous speech recognition output. We propose pronunciation-assisted sub-word modeling (PASM), a sub-word extraction method that leverages the pronunciation information of a word. Experiments show that the proposed method can greatly improve upon the characterbased baseline, and also outperform commonly used byte-pair encoding methods.
|Publication status||Published - 2018 Nov 10|
- End-to-end models
- Speech recognition
- Sub-word modeling
ASJC Scopus subject areas