Efficient Multilingual Multi-modal Pre-training through Triple Contrastive Loss

Youhan Lee, KyungTae Lim, Woonhyuk Baek, Byungseok Roh, Saehoon Kim


Abstract
Learning visual and textual representations in the shared space from web-scale image-text pairs improves the performance of diverse vision-and-language tasks, as well as modality-specific tasks. Many attempts in this framework have been made to connect English-only texts and images, and only a few works have been proposed to extend this framework in multilingual settings with the help of many translation pairs. In this multilingual approach, a typical setup is to use pairs of (image and English-text) and translation pairs. The major limitation of this approach is that the learning signal of aligning visual representation with under-resourced language representation is not strong, achieving a sub-optimal performance of vision-and-language tasks. In this work, we propose a simple yet effective enhancement scheme for previous multilingual multi-modal representation methods by using a limited number of pairs of images and non-English texts. In specific, our scheme fine-tunes a pre-trained multilingual model by minimizing a triplet contrastive loss on triplets of image and two different language texts with the same meaning, improving the connection between images and non-English texts. Experiments confirm that our enhancement strategy achieves performance gains in image-text retrieval, zero-shot image classification, and sentence embedding tasks.
Anthology ID:
2022.coling-1.504
Volume:
Proceedings of the 29th International Conference on Computational Linguistics
Month:
October
Year:
2022
Address:
Gyeongju, Republic of Korea
Venue:
COLING
SIG:
Publisher:
International Committee on Computational Linguistics
Note:
Pages:
5730–5744
Language:
URL:
https://aclanthology.org/2022.coling-1.504
DOI:
Bibkey:
Cite (ACL):
Youhan Lee, KyungTae Lim, Woonhyuk Baek, Byungseok Roh, and Saehoon Kim. 2022. Efficient Multilingual Multi-modal Pre-training through Triple Contrastive Loss. In Proceedings of the 29th International Conference on Computational Linguistics, pages 5730–5744, Gyeongju, Republic of Korea. International Committee on Computational Linguistics.
Cite (Informal):
Efficient Multilingual Multi-modal Pre-training through Triple Contrastive Loss (Lee et al., COLING 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.coling-1.504.pdf
Data
CCMatrixCIFAR-100COCOFlickr30kFood-101ImageNet