Segmental Recurrent Neural Networks for End-to-end Speech Recognition
This work addresses speech recognition by enabling self-contained, end-to-end training without external features, though it is incremental as it builds on existing CRF and RNN methods.
The authors tackled end-to-end acoustic modeling for speech recognition by combining a segmental conditional random field with a recurrent neural network, achieving a phone error rate of 17.3% on the TIMIT dataset without a language model.
We study the segmental recurrent neural network for end-to-end acoustic modelling. This model connects the segmental conditional random field (CRF) with a recurrent neural network (RNN) used for feature extraction. Compared to most previous CRF-based acoustic models, it does not rely on an external system to provide features or segmentation boundaries. Instead, this model marginalises out all the possible segmentations, and features are extracted from the RNN trained together with the segmental CRF. In essence, this model is self-contained and can be trained end-to-end. In this paper, we discuss practical training and decoding issues as well as the method to speed up the training in the context of speech recognition. We performed experiments on the TIMIT dataset. We achieved 17.3 phone error rate (PER) from the first-pass decoding --- the best reported result using CRFs, despite the fact that we only used a zeroth-order CRF and without using any language model.