Token-wise Training for Attention Based End-to-end Speech Recognition

By Peidong Wang, Jia Cui, Chao Weng, Dong Yu. 2019

5 citations recorded by the source corpus.

In attention based end-to-end (A-E2E) speech recognition systems, the dependency between output tokens is typically formulated as an input-output mapping in decoder. Due to such dependency, decoding errors can easily propagate along output sequence. In this paper, we propose a token-wise training (TWT) method for A-E2E models. The new method is flexible and can be combined with a variety of loss functions. Applying TWT to multiple hypotheses, we propose a novel TWT in beam (TWTiB) training scheme. Trained on the benchmark Switchboard (SWBD) 300h corpus, TWTiB outperforms the previous best training scheme on the SWBD evaluation subset.

Subfields: Artificial Intelligence, Speech & Audio, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Ron J. Weiss
2. Tara N. Sainath
3. Jinyu Li
4. Peidong Wang
5. Dong Yu
6. Yu Zhang
7. Yonghui Wu
8. Bo Li

In Speech & Audio

1. Tara N. Sainath
2. Ron J. Weiss
3. Peidong Wang
4. Jinyu Li
5. Yu Zhang
6. Yonghui Wu
7. Dong Yu
8. Bo Li

In Signal Processing

1. Ron J. Weiss
2. Tara N. Sainath
3. Peidong Wang
4. Jinyu Li
5. Dong Yu
6. Yu Zhang
7. Bo Li
8. Yonghui Wu

Citations in the corpus

• Multitask Training with Text Data for End-to-End Speech Recognition (2020, background): “Various model architectures [6, 7, 8, 9, 10] and training schemes [11, 12, 13] were proposed.”