English

L2RS: A Learning-to-Rescore Mechanism for Automatic Speech Recognition

Computation and Language 2019-10-28 v1 Sound Audio and Speech Processing

Abstract

Modern Automatic Speech Recognition (ASR) systems primarily rely on scores from an Acoustic Model (AM) and a Language Model (LM) to rescore the N-best lists. With the abundance of recent natural language processing advances, the information utilized by current ASR for evaluating the linguistic and semantic legitimacy of the N-best hypotheses is rather limited. In this paper, we propose a novel Learning-to-Rescore (L2RS) mechanism, which is specialized for utilizing a wide range of textual information from the state-of-the-art NLP models and automatically deciding their weights to rescore the N-best lists for ASR systems. Specifically, we incorporate features including BERT sentence embedding, topic vector, and perplexity scores produced by n-gram LM, topic modeling LM, BERT LM and RNNLM to train a rescoring model. We conduct extensive experiments based on a public dataset, and experimental results show that L2RS outperforms not only traditional rescoring methods but also its deep neural network counterparts by a substantial improvement of 20.67% in terms of NDCG@10. L2RS paves the way for developing more effective rescoring models for ASR.

Keywords

Cite

@article{arxiv.1910.11496,
  title  = {L2RS: A Learning-to-Rescore Mechanism for Automatic Speech Recognition},
  author = {Yuanfeng Song and Di Jiang and Xuefang Zhao and Qian Xu and Raymond Chi-Wing Wong and Lixin Fan and Qiang Yang},
  journal= {arXiv preprint arXiv:1910.11496},
  year   = {2019}
}

Comments

5 pages, 3 figures

R2 v1 2026-06-23T11:54:28.540Z