On Using SpecAugment for End-to-End Speech Translation
arXiv:1911.08876
Abstract
This work investigates a simple data augmentation technique, SpecAugment, for end-to-end speech translation. SpecAugment is a low-cost implementation method applied directly to the audio input features and it consists of masking blocks of frequency channels, and/or time steps. We apply SpecAugment on end-to-end speech translation tasks and achieve up to +2.2\% \BLEU on LibriSpeech Audiobooks En->Fr and +1.2% on IWSLT TED-talks En->De by alleviating overfitting to some extent. We also examine the effectiveness of the method in a variety of data scenarios and show that the method also leads to significant improvements in various data conditions irrespective of the amount of training data.
8 pages, International Workshop on Spoken Language Translation (IWSLT), Hong Kong, China, November 2019
References in corpus (7)
- Sequence to Sequence Learning with Neural Networks
- SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
- Regularizing Neural Networks by Penalizing Confident Output Distributions
- Improved training of end-to-end attention models for speech recognition
- Listen and Translate: A Proof of Concept for End-to-End Speech-to-Text Translation
- End-to-End Speech Translation with Knowledge Distillation
- Towards Unsupervised Speech-to-Text Translation
Cited by in corpus (6)
- Bridging the Modality Gap for Speech-to-Text Translation
- MAM: Masked Acoustic Modeling for End-to-End Speech-to-Text Translation
- Learning Shared Semantic Space for Speech-to-Text Translation
- On Target Segmentation for Direct Speech Translation
- The RWTH ASR System for TED-LIUM Release 2: Improving Hybrid HMM with SpecAugment
- A Technical Report: BUT Speech Translation Systems