Adversarial Multilingual Training for Low-Resource Speech Recognition

Jiangyan Yi, Jianhua Tao, Zhengqi Wen, Ye BaiICASSP 2018 · OpenReview

Abstract

This paper proposes an adversarial multilingual training to train bottleneck (BN) networks for the target language. A parallel shared-exclusive model is also proposed to train the BN network. Adversarial training is used to ensure that the shared layers can learn language-invariant features. Experiments are conducted on IARPA Babel datasets. The results show that the proposed adversarial multilingual BN model outperforms the baseline BN model by up to 8.9% relative word error rate (WER) reduction. The results also show that the proposed parallel shared-exclusive model achieves up to 1.7% relative WER reduction when compared with the stacked share-exclusive model.
BibTeX
@inproceedings{yi2018adversarial,
  author = {Jiangyan Yi and Jianhua Tao and Zhengqi Wen and Ye Bai},
  title = {{Adversarial Multilingual Training for Low-Resource Speech Recognition}},
  booktitle = {IEEE International Conference on Acoustics, Speech and Signal Processing},
  year = {2018},
  url = {https://openreview.net/forum?id=bfGbuSErhVK},
}

Loading related content...

Figure