Make More of Your Data: Minimal Effort Data Augmentation for Automatic Speech Recognition and Translation
This incremental work addresses the need for cost-effective performance gains in speech processing for researchers and practitioners, showing improvements with simple augmentation.
The paper tackles the problem of improving automatic speech recognition and translation models by proposing a minimal-effort data augmentation method based on concatenating existing training examples, resulting in significant WER reductions such as from 2.83 to 2.55 on LibriSpeech test-clean and up to 0.9 WER on CoVoST-2 for non-English languages.
Data augmentation is a technique to generate new training data based on existing data. We evaluate the simple and cost-effective method of concatenating the original data examples to build new training instances. Continued training with such augmented data is able to improve off-the-shelf Transformer and Conformer models that were optimized on the original data only. We demonstrate considerable improvements on the LibriSpeech-960h test sets (WER 2.83 and 6.87 for test-clean and test-other), which carry over to models combined with shallow fusion (WER 2.55 and 6.27). Our method of continued training also leads to improvements of up to 0.9 WER on the ASR part of CoVoST-2 for four non English languages, and we observe that the gains are highly dependent on the size of the original training data. We compare different concatenation strategies and found that our method does not need speaker information to achieve its improvements. Finally, we demonstrate on two datasets that our methods also works for speech translation tasks.