Recurrent DNNs and its Ensembles on the TIMIT Phone Recognition Task
This work addresses improving acoustic model accuracy for speech recognition, particularly in low-resource scenarios, but it is incremental as it builds on existing recurrent DNN methods.
The paper tackled phone recognition on the TIMIT benchmark by investigating recurrent DNNs with regularization techniques like dropout and zoneout, and found that an ensemble of these models achieved a state-of-the-art average phone error rate of 14.84%.
In this paper, we have investigated recurrent deep neural networks (DNNs) in combination with regularization techniques as dropout, zoneout, and regularization post-layer. As a benchmark, we chose the TIMIT phone recognition task due to its popularity and broad availability in the community. It also simulates a low-resource scenario that is helpful in minor languages. Also, we prefer the phone recognition task because it is much more sensitive to an acoustic model quality than a large vocabulary continuous speech recognition task. In recent years, recurrent DNNs pushed the error rates in automatic speech recognition down. But, there was no clear winner in proposed architectures. The dropout was used as the regularization technique in most cases, but combination with other regularization techniques together with model ensembles was omitted. However, just an ensemble of recurrent DNNs performed best and achieved an average phone error rate from 10 experiments 14.84 % (minimum 14.69 %) on core test set that is slightly lower then the best-published PER to date, according to our knowledge. Finally, in contrast of the most papers, we published the open-source scripts to easily replicate the results and to help continue the development.