Speech-Transformer: A No-Recurrence Sequence-to-Sequence Model for Speech Recognition

Recurrent sequence-to-sequence models using encoder-decoder architecture have made great progress in speech recognition task. However, they suffer from the drawback of slow training speed because the internal recurrence limits the training parallelization. In this paper, we present the Speech-Transformer, a no-recurrence sequence-to-sequence model entirely relies on attention mechanisms to learn the positional dependencies, which can be trained faster with more efficiency. We also propose a 2D-Attention mechanism, which can jointly attend to the time and frequency axes of the 2-dimensional speech inputs, thus providing more expressive representations for the Speech-Transformer. Evaluated on the Wall Street Journal (WSJ) speech recognition dataset, our best model achieves competitive word error rate (WER) of 10.9%, while the whole training process only takes 1.2 days on 1 GPU, significantly faster than the published results of recurrent sequence-to-sequence models.

[1]  Geoffrey Zweig,et al.  Exploring multidimensional lstms for large vocabulary ASR , 2016, 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[2]  Satoshi Nakamura,et al.  Local Monotonic Attention Mechanism for End-to-End Speech Recognition , 2017, ArXiv.

[3]  Navdeep Jaitly,et al.  Towards Better Decoding and Language Model Integration in Sequence to Sequence Models , 2016, INTERSPEECH.

[4]  Lawrence R. Rabiner,et al.  A tutorial on hidden Markov models and selected applications in speech recognition , 1989, Proc. IEEE.

[5]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[6]  Alex Graves,et al.  Sequence Transduction with Recurrent Neural Networks , 2012, ArXiv.

[7]  Quoc V. Le,et al.  Listen, attend and spell: A neural network for large vocabulary conversational speech recognition , 2015, 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[8]  Geoffrey E. Hinton,et al.  Speech recognition with deep recurrent neural networks , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[9]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[10]  Navdeep Jaitly,et al.  Towards End-To-End Speech Recognition with Recurrent Neural Networks , 2014, ICML.

[11]  Dong Yu,et al.  Recent progresses in deep learning based acoustic models , 2017, IEEE/CAA Journal of Automatica Sinica.

[12]  Andrew W. Senior,et al.  Long short-term memory recurrent neural network architectures for large scale acoustic modeling , 2014, INTERSPEECH.

[13]  Martín Abadi,et al.  TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems , 2016, ArXiv.

[14]  Yoshua Bengio,et al.  Attention-Based Models for Speech Recognition , 2015, NIPS.

[15]  Yoshua Bengio,et al.  Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling , 2014, ArXiv.

[16]  Hairong Liu,et al.  Exploring neural transducers for end-to-end speech recognition , 2017, 2017 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU).

[17]  Yu Zhang,et al.  Latent Sequence Decompositions , 2016, ICLR.

[18]  Yu Zhang,et al.  Very deep convolutional networks for end-to-end speech recognition , 2016, 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[19]  Yoshua Bengio,et al.  End-to-end attention-based large vocabulary speech recognition , 2015, 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[20]  Shinji Watanabe,et al.  Joint CTC-attention based end-to-end speech recognition using multi-task learning , 2016, 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[21]  A. Algorithms.  Online and Linear-Time Attention by Enforcing Monotonic Alignments , 2017 .

[22]  George Kurian,et al.  Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation , 2016, ArXiv.