ASAPP-ASR: Multistream CNN and Self-Attentive SRU for SOTA Speech Recognition

Jing Pan, Joshua Shapiro, Jeremy Wohlwend, Kyu J. Han, Tao Lei, Tao Ma


In this paper we present state-of-the-art (SOTA) performance on the LibriSpeech corpus with two novel neural network architectures, a multistream CNN for acoustic modeling and a self-attentive simple recurrent unit (SRU) for language modeling. In the hybrid ASR framework, the multistream CNN acoustic model processes an input of speech frames in multiple parallel pipelines where each stream has a unique dilation rate for diversity. Trained with the SpecAugment data augmentation method, it achieves relative word error rate (WER) improvements of 4% on test-clean and 14% on test-other. We further improve the performance via N-best rescoring using a 24-layer self-attentive SRU language model, achieving WERs of 1.75% on test-clean and 4.46% on test-other.


 DOI: 10.21437/Interspeech.2020-2947

Cite as: Pan, J., Shapiro, J., Wohlwend, J., Han, K.J., Lei, T., Ma, T. (2020) ASAPP-ASR: Multistream CNN and Self-Attentive SRU for SOTA Speech Recognition. Proc. Interspeech 2020, 16-20, DOI: 10.21437/Interspeech.2020-2947.


@inproceedings{Pan2020,
  author={Jing Pan and Joshua Shapiro and Jeremy Wohlwend and Kyu J. Han and Tao Lei and Tao Ma},
  title={{ASAPP-ASR: Multistream CNN and Self-Attentive SRU for SOTA Speech Recognition}},
  year=2020,
  booktitle={Proc. Interspeech 2020},
  pages={16--20},
  doi={10.21437/Interspeech.2020-2947},
  url={http://dx.doi.org/10.21437/Interspeech.2020-2947}
}