Semi-Supervised Speech Recognition via Local Prior Matching

For sequence transduction tasks like speech recognition, a strong structured prior model encodes rich information about the target space, implicitly ruling out invalid sequences by assigning them low probability. In this work, we propose local prior matching (LPM), a semi-supervised objective that distills knowledge from a strong prior (e.g. a language model) to provide learning signal to a discriminative model trained on unlabeled speech. We demonstrate that LPM is theoretically well-motivated, simple to implement, and superior to existing knowledge distillation techniques under comparable settings. Starting from a baseline trained on 100 hours of labeled speech, with an additional 360 hours of unlabeled data, LPM recovers 54% and 73% of the word error rate on clean and noisy test sets relative to a fully supervised model on the same data.

[1]  Tara N. Sainath,et al.  State-of-the-Art Speech Recognition with Sequence-to-Sequence Models , 2017, 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[2]  Hung-yi Lee,et al.  Gate Activation Signal Analysis for Gated Recurrent Neural Networks and its Correlation with Phoneme Boundaries , 2017, INTERSPEECH.

[3]  P. Kuhl Early language acquisition: cracking the speech code , 2004, Nature Reviews Neuroscience.

[4]  Hao Tang,et al.  Unsupervised Adaptation with Interpretable Disentangled Representations for Distant Conversational Speech Recognition , 2018, INTERSPEECH.

[5]  Yann Dauphin,et al.  Language Modeling with Gated Convolutional Networks , 2016, ICML.

[6]  Ramón Fernández Astudillo,et al.  Self-supervised Sequence-to-sequence ASR using Unpaired Speech and Text , 2019, INTERSPEECH.

[7]  Sanjeev Khudanpur,et al.  Semi-Supervised Training of Acoustic Models Using Lattice-Free MMI , 2018, 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[8]  Erich Elsen,et al.  Deep Speech: Scaling up end-to-end speech recognition , 2014, ArXiv.

[9]  Biing-Hwang Juang,et al.  Adversarial Teacher-Student Learning for Unsupervised Domain Adaptation , 2018, 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[10]  Lukás Burget,et al.  Semi-Supervised DNN Training with Word Selection for ASR , 2017, INTERSPEECH.

[11]  Ben Taskar,et al.  Posterior Regularization for Structured Latent Variable Models , 2010, J. Mach. Learn. Res..

[12]  Awni Hannun,et al.  Self-Training for End-to-End Speech Recognition , 2020, ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[13]  James R. Glass,et al.  Extracting Domain Invariant Features by Unsupervised Learning for Robust Automatic Speech Recognition , 2018, 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[14]  Nicolas Usunier,et al.  Fully Convolutional Speech Recognition , 2018, ArXiv.

[15]  Jonathan Le Roux,et al.  Discriminative Training for Large-Vocabulary Speech Recognition Using Minimum Classification Error , 2007, IEEE Transactions on Audio, Speech, and Language Processing.

[16]  Jordi Adell,et al.  Towards phone segmentation for concatenative speech synthesis , 2004, SSW.

[17]  Armand Joulin,et al.  Libri-Light: A Benchmark for ASR with Limited or No Supervision , 2020, ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[18]  Lin-Shan Lee,et al.  Adversarial Training of End-to-end Speech Recognition Using a Criticizing Language Model , 2018, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[19]  Emmanuel Dupoux,et al.  Cognitive science in the era of artificial intelligence: A roadmap for reverse-engineering the infant language-learner , 2016, Cognition.

[20]  Kristin Precoda Non-Mainstream Languages and Speech Recognition: Some Challenges , 2013 .

[21]  S. Freytag Knowledge Of Language Its Nature Origin And Use , 2016 .

[22]  Wei Ping,et al.  Non-Autoregressive Neural Text-to-Speech , 2020, ICML.

[23]  Zhao Song,et al.  Parallel Neural Text-to-Speech , 2019, ArXiv.

[24]  Yifan Gong,et al.  Unsupervised adaptation with domain separation networks for robust speech recognition , 2017, 2017 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU).

[25]  Yu Zhang,et al.  Unsupervised domain adaptation for robust speech recognition via variational autoencoder-based data augmentation , 2017, 2017 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU).

[26]  Gabriel Synnaeve,et al.  A Fully Differentiable Beam Search Decoder , 2019, ICML.

[27]  Brian Kingsbury,et al.  Knowledge distillation across ensembles of multilingual models for low-resource languages , 2017, 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

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

[29]  Ronan Collobert,et al.  Sequence-to-Sequence Speech Recognition with Time-Depth Separable Convolutions , 2019, INTERSPEECH.

[30]  Sanjeev Khudanpur,et al.  Librispeech: An ASR corpus based on public domain audio books , 2015, 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[31]  Yifan Gong,et al.  Attentive Adversarial Learning for Domain-invariant Training , 2019, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[32]  Satoshi Nakamura,et al.  End-to-end Feedback Loss in Speech Chain Framework via Straight-through Estimator , 2019, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[33]  Odette Scharenborg,et al.  Unsupervised speech segmentation: an analysis of the hypothesized phone boundaries. , 2010, The Journal of the Acoustical Society of America.

[34]  Taku Kudo,et al.  Subword Regularization: Improving Neural Network Translation Models with Multiple Subword Candidates , 2018, ACL.

[35]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[36]  Taku Kudo,et al.  SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing , 2018, EMNLP.

[37]  Chong Wang,et al.  Deep Speech 2 : End-to-End Speech Recognition in English and Mandarin , 2015, ICML.

[38]  Rico Sennrich,et al.  Improving Neural Machine Translation Models with Monolingual Data , 2015, ACL.

[39]  拓海 杉山,et al.  “Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告 , 2017 .

[40]  Tomoharu Iwata,et al.  Semi-Supervised End-to-End Speech Recognition , 2018, INTERSPEECH.

[41]  Bowen Zhou,et al.  Abstractive Text Summarization using Sequence-to-sequence RNNs and Beyond , 2016, CoNLL.

[42]  Ronan Collobert,et al.  Wav2Letter++: A Fast Open-source Speech Recognition System , 2018, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[43]  Jonathan Le Roux,et al.  Cycle-consistency Training for End-to-end Speech Recognition , 2018, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[44]  Lukás Burget,et al.  Sequence-discriminative training of deep neural networks , 2013, INTERSPEECH.

[45]  Gabriel Synnaeve,et al.  Wav2Letter++: A Fast Open-source Speech Recognition System , 2018, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

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

[47]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[48]  Sree Hari Krishnan Parthasarathi,et al.  Lessons from Building Acoustic Models with a Million Hours of Speech , 2019, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[49]  Mitch Weintraub,et al.  Acoustic Modeling for Google Home , 2017, INTERSPEECH.

[50]  James R. Glass Towards unsupervised speech processing , 2012, 2012 11th International Conference on Information Science, Signal Processing and their Applications (ISSPA).

[51]  Tara N. Sainath,et al.  Semi-supervised Training for End-to-end Models via Weak Distillation , 2019, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[52]  Satoshi Nakamura,et al.  Listening while speaking: Speech chain by deep learning , 2017, 2017 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU).

[53]  Tomoki Toda,et al.  Back-Translation-Style Data Augmentation for end-to-end ASR , 2018, 2018 IEEE Spoken Language Technology Workshop (SLT).

[54]  Yanning Zhang,et al.  An unsupervised deep domain adaptation approach for robust speech recognition , 2017, Neurocomputing.

[55]  James R. Glass,et al.  Combining End-to-End and Adversarial Training for Low-Resource Speech Recognition , 2018, 2018 IEEE Spoken Language Technology Workshop (SLT).

[56]  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).