End-to-End Phoneme Recognition using Models from Semantic Image Segmentation

Wei Gao, Ahmad Hashemi-Sakhtsari, Mark D. McDonnellIJCNN 2020 · OpenReview

Abstract

We train fully convolutional neural networks with no recurrent layers for the end-to-end phoneme recognition task, using the Connectionist Temporal Classification (CTC) loss function. The adopted network, U-Net, was introduced initially for semantic image segmentation tasks, and is often applied to segmenting features in medical imaging and remote sensing. The similarities between CTC-based automatic speech recognition and semantic segmentation problems are discussed. We extend the encoder-decoder architecture of U-Net and show it is capable of good performance in the acoustic modelling of a speech recognition system. We investigate the importance of the concatenation step in the design of U-net, and report results using the core test set of the TIMIT corpus.
BibTeX
@inproceedings{gao2020end,
  author = {Wei Gao and Ahmad Hashemi-Sakhtsari and Mark D. McDonnell},
  title = {{End-to-End Phoneme Recognition using Models from Semantic Image Segmentation}},
  booktitle = {IJCNN},
  year = {2020},
  url = {https://openreview.net/forum?id=qQw63tCIZ6},
}

Loading related content...

Figure