English

Word-level Speech Recognition with a Letter to Word Encoder

Computation and Language 2020-07-16 v2 Sound Audio and Speech Processing

Abstract

We propose a direct-to-word sequence model which uses a word network to learn word embeddings from letters. The word network can be integrated seamlessly with arbitrary sequence models including Connectionist Temporal Classification and encoder-decoder models with attention. We show our direct-to-word model can achieve word error rate gains over sub-word level models for speech recognition. We also show that our direct-to-word approach retains the ability to predict words not seen at training time without any retraining. Finally, we demonstrate that a word-level model can use a larger stride than a sub-word level model while maintaining accuracy. This makes the model more efficient both for training and inference.

Keywords

Cite

@article{arxiv.1906.04323,
  title  = {Word-level Speech Recognition with a Letter to Word Encoder},
  author = {Ronan Collobert and Awni Hannun and Gabriel Synnaeve},
  journal= {arXiv preprint arXiv:1906.04323},
  year   = {2020}
}

Comments

ICML 2020

R2 v1 2026-06-23T09:49:36.737Z