Normalized Contrastive Learning for Text-Video Retrieval

Yookoon Park, Mahmoud Azab, Seungwhan Moon, Bo Xiong, Florian Metze, Gourab Kundu, Kirmani Ahmed


Abstract
Cross-modal contrastive learning has led the recent advances in multimodal retrieval with its simplicity and effectiveness. In this work, however, we reveal that cross-modal contrastive learning suffers from incorrect normalization of the sum retrieval probabilities of each text or video instance. Specifically, we show that many test instances are either over- or under-represented during retrieval, significantly hurting the retrieval performance. To address this problem, we propose Normalized Contrastive Learning (NCL) which utilizes the Sinkhorn-Knopp algorithm to compute the instance-wise biases that properly normalize the sum retrieval probabilities of each instance so that every text and video instance is fairly represented during cross-modal retrieval. Empirical study shows that NCL brings consistent and significant gains in text-video retrieval on different model architectures, with new state-of-the-art multimodal retrieval metrics on the ActivityNet, MSVD, and MSR-VTT datasets without any architecture engineering.
Anthology ID:
2022.emnlp-main.17
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
248–260
Language:
URL:
https://aclanthology.org/2022.emnlp-main.17
DOI:
10.18653/v1/2022.emnlp-main.17
Bibkey:
Cite (ACL):
Yookoon Park, Mahmoud Azab, Seungwhan Moon, Bo Xiong, Florian Metze, Gourab Kundu, and Kirmani Ahmed. 2022. Normalized Contrastive Learning for Text-Video Retrieval. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 248–260, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
Normalized Contrastive Learning for Text-Video Retrieval (Park et al., EMNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.emnlp-main.17.pdf