تقرير
End-to-End Speech Translation with Pre-trained Models and Adapters: UPC at IWSLT 2021
| العنوان: | End-to-End Speech Translation with Pre-trained Models and Adapters: UPC at IWSLT 2021 |
|---|---|
| المؤلفون: | Gállego, Gerard I., Tsiamas, Ioannis, Escolano, Carlos, Fonollosa, José A. R., Costa-jussà, Marta R. |
| سنة النشر: | 2021 |
| المجموعة: | Computer Science |
| مصطلحات موضوعية: | Computer Science - Computation and Language |
| الوصف: | This paper describes the submission to the IWSLT 2021 offline speech translation task by the UPC Machine Translation group. The task consists of building a system capable of translating English audio recordings extracted from TED talks into German text. Submitted systems can be either cascade or end-to-end and use a custom or given segmentation. Our submission is an end-to-end speech translation system, which combines pre-trained models (Wav2Vec 2.0 and mBART) with coupling modules between the encoder and decoder, and uses an efficient fine-tuning technique, which trains only 20% of its total parameters. We show that adding an Adapter to the system and pre-training it, can increase the convergence speed and the final result, with which we achieve a BLEU score of 27.3 on the MuST-C test set. Our final model is an ensemble that obtains 28.22 BLEU score on the same set. Our submission also uses a custom segmentation algorithm that employs pre-trained Wav2Vec 2.0 for identifying periods of untranscribable text and can bring improvements of 2.5 to 3 BLEU score on the IWSLT 2019 test set, as compared to the result with the given segmentation. Comment: Submitted to IWSLT 2021; changed the title and added submission results |
| نوع الوثيقة: | Working Paper |
| URL الوصول: | http://arxiv.org/abs/2105.04512 |
| رقم الأكسشن: | edsarx.2105.04512 |
| قاعدة البيانات: | arXiv |
| الوصف غير متاح. |