Bayesian Neural Network Language Modeling for Speech Recognition
Publication in refereed journal
CUHK Authors
Full Text
Digital Object Identifier (DOI) DOI for CUHK Users |
Times Cited
Altmetrics Information
.
Other information
AbstractState-of-the-art neural network language models (NNLMs) represented by long short term memory recurrent neural networks (LSTM-RNNs) and Transformers are becoming highly complex. They are prone to overfitting and poor generalization when given limited training data. To this end, an overarching full Bayesian learning framework encompassing three methods is proposed in this paper to account for the underlying uncertainty in LSTM-RNN and Transformer LMs. The uncertainty over their model parameters, choice of neural activations and hidden output representations are modeled using Bayesian, Gaussian Process and variational LSTM-RNN or Transformer LMs respectively. Efficient inference approaches were used to automatically select the optimal network internal components to be Bayesian learned using neural architecture search. A minimal number of Monte Carlo parameter samples as low as one was also used. These allow the computational costs incurred in Bayesian NNLM training and evaluation to be minimized. Experiments are conducted on two tasks: AMI meeting transcription and Oxford-BBC LipReading Sentences 2 (LRS2) overlapped speech recognition using state-of-the-art LF-MMI trained factored TDNN systems featuring data augmentation, speaker adaptation and audio-visual multi-channel beamforming for overlapped speech. Consistent performance improvements over the baseline LSTM-RNN and Transformer LMs with point estimated model parameters and drop-out regularization were obtained across both tasks in terms of perplexity and word error rate (WER). In particular, on the LRS2 data, statistically signi cant WER reductions up to 1.3% and 1.2% absolute (12.1% and 11.3% relative) were obtained over the baseline LSTM-RNN and Transformer LMs respectively after model combination between Bayesian NNLMs and their respective baselines.
Acceptance Date25/08/2022
All Author(s) ListBoyang Xue, Shoukang Hu, Junhao Xu, Mengzhe Geng, Xunying Liu, Helen Meng
Journal nameIEEE/ACM Transactions on Audio, Speech and Language Processing
Year2022
Month9
Volume Number30
PublisherIEEE
Pages2900 - 2917
ISSN2329-9290
LanguagesEnglish-United States
Keywordsneural language models, Bayesian learning, model uncertainty, neural architecture search, speech recognition
