Dealing with training and test segmentation mismatch: FBK@IWSLT2021

Sara Papi, Marco Gaido, Matteo Negri, Marco Turchi


Abstract
This paper describes FBK’s system submission to the IWSLT 2021 Offline Speech Translation task. We participated with a direct model, which is a Transformer-based architecture trained to translate English speech audio data into German texts. The training pipeline is characterized by knowledge distillation and a two-step fine-tuning procedure. Both knowledge distillation and the first fine-tuning step are carried out on manually segmented real and synthetic data, the latter being generated with an MT system trained on the available corpora. Differently, the second fine-tuning step is carried out on a random segmentation of the MuST-C v2 En-De dataset. Its main goal is to reduce the performance drops occurring when a speech translation model trained on manually segmented data (i.e. an ideal, sentence-like segmentation) is evaluated on automatically segmented audio (i.e. actual, more realistic testing conditions). For the same purpose, a custom hybrid segmentation procedure that accounts for both audio content (pauses) and for the length of the produced segments is applied to the test data before passing them to the system. At inference time, we compared this procedure with a baseline segmentation method based on Voice Activity Detection (VAD). Our results indicate the effectiveness of the proposed hybrid approach, shown by a reduction of the gap with manual segmentation from 8.3 to 1.4 BLEU points.
Anthology ID:
2021.iwslt-1.8
Volume:
Proceedings of the 18th International Conference on Spoken Language Translation (IWSLT 2021)
Month:
August
Year:
2021
Address:
Bangkok, Thailand (online)
Editors:
Marcello Federico, Alex Waibel, Marta R. Costa-jussà, Jan Niehues, Sebastian Stuker, Elizabeth Salesky
Venue:
IWSLT
SIG:
SIGSLT
Publisher:
Association for Computational Linguistics
Note:
Pages:
84–91
Language:
URL:
https://aclanthology.org/2021.iwslt-1.8
DOI:
10.18653/v1/2021.iwslt-1.8
Bibkey:
Cite (ACL):
Sara Papi, Marco Gaido, Matteo Negri, and Marco Turchi. 2021. Dealing with training and test segmentation mismatch: FBK@IWSLT2021. In Proceedings of the 18th International Conference on Spoken Language Translation (IWSLT 2021), pages 84–91, Bangkok, Thailand (online). Association for Computational Linguistics.
Cite (Informal):
Dealing with training and test segmentation mismatch: FBK@IWSLT2021 (Papi et al., IWSLT 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.iwslt-1.8.pdf