LIUM-CVC submissions for WMT18 multimodal translation task
File(s)LIUM-CVC-Submissions-for-WMT18.pdf (336.91 KB)
Working paper
Author(s)
Type
Working Paper
Abstract
This paper describes the multimodal Neural Machine Translation systems
developed by LIUM and CVC for WMT18 Shared Task on Multimodal Translation. This
year we propose several modifications to our previous multimodal attention
architecture in order to better integrate convolutional features and refine
them using encoder-side information. Our final constrained submissions ranked
first for English-French and second for English-German language pairs among the
constrained submissions according to the automatic evaluation metric METEOR.
developed by LIUM and CVC for WMT18 Shared Task on Multimodal Translation. This
year we propose several modifications to our previous multimodal attention
architecture in order to better integrate convolutional features and refine
them using encoder-side information. Our final constrained submissions ranked
first for English-French and second for English-German language pairs among the
constrained submissions according to the automatic evaluation metric METEOR.
Date Issued
2018-09-01
Citation
2018
Publisher
arxiv
Copyright Statement
© 2018 The Authors.
Identifier
http://arxiv.org/abs/1809.00151v1
Subjects
cs.CL
cs.CL
Notes
WMT2018
Publication Status
Published
OA Location
https://www.aclweb.org/anthology/W18-6438.pdf