ON-TRAC Consortium End-to-End Speech Translation Systems for the IWSLT 2019 Shared Task

This paper describes the ON-TRAC Consortium translation systems developed for\nthe end-to-end model task of IWSLT Evaluation 2019 for the\nEnglish-to-Portuguese language pair. ON-TRAC Consortium is composed of\nresearchers from three French academic laboratories: LIA (Avignon\nUniversit\\'e), LIG (Universit\\'e Grenoble Alpes), and LIUM (Le Mans\nUniversit\\'e). A single end-to-end model built as a neural encoder-decoder\narchitecture with attention mechanism was used for two primary submissions\ncorresponding to the two EN-PT evaluations sets: (1) TED (MuST-C) and (2) How2.\nIn this paper, we notably investigate impact of pooling heterogeneous corpora\nfor training, impact of target tokenization (characters or BPEs), impact of\nspeech input segmentation and we also compare our best end-to-end model (BLEU\nof 26.91 on MuST-C and 43.82 on How2 validation sets) to a pipeline (ASR+MT)\napproach.\n

Paper

Similar papers

© 2026 NYSGPT2525 LLC