Lessons learned in multilingual grounded language learning
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Recent work has shown how to learn bettervisual-semantic embeddings by leveraging imagedescriptions in more than one language.Here, we investigate in detail which conditionsaffect the performance of this type of grounded language learning model. We show that multilingual training improves over bilingual training, and that low-resource languages benefit from training with higher-resource languages. We demonstrate that a multilingual model can be trained equally well on either translations or comparable sentence pairs, and that annotating the same set of images in multiple language enables further improvements via an additional caption-caption ranking objective
Original language | English |
---|---|
Title of host publication | Proceedings of the 22nd Conference on Computational Natural Language Learning |
Number of pages | 11 |
Publisher | Association for Computational Linguistics |
Publication date | 2018 |
Pages | 402-412 |
ISBN (Print) | 978-1-948087-72-8 |
Publication status | Published - 2018 |
Event | 22nd Conference on Computational Natural Language Learning (CoNLL 2018) - Brussels, Belgium Duration: 31 Oct 2018 → 1 Nov 2018 |
Conference
Conference | 22nd Conference on Computational Natural Language Learning (CoNLL 2018) |
---|---|
Land | Belgium |
By | Brussels |
Periode | 31/10/2018 → 01/11/2018 |
ID: 230797458