Updating CLIP to Prefer Descriptions Over Captions

Amir Zur, Elisa Kreiss, Karel D’Oosterlinck, Christopher Potts, Atticus Geiger


Abstract
Although CLIPScore is a powerful generic metric that captures the similarity between a text and an image, it fails to distinguish between a caption that is meant to complement the information in an image and a description that is meant to replace an image entirely, e.g., for accessibility. We address this shortcoming by updating the CLIP model with the Concadia dataset to assign higher scores to descriptions than captions using parameter efficient fine-tuning and a loss objective derived from work on causal interpretability. This model correlates with the judgements of blind and low-vision people while preserving transfer capabilities and has interpretable structure that sheds light on the caption–description distinction.
Anthology ID:
2024.emnlp-main.1125
Volume:
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
Month:
November
Year:
2024
Address:
Miami, Florida, USA
Editors:
Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
20178–20187
Language:
URL:
https://aclanthology.org/2024.emnlp-main.1125
DOI:
Bibkey:
Cite (ACL):
Amir Zur, Elisa Kreiss, Karel D’Oosterlinck, Christopher Potts, and Atticus Geiger. 2024. Updating CLIP to Prefer Descriptions Over Captions. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 20178–20187, Miami, Florida, USA. Association for Computational Linguistics.
Cite (Informal):
Updating CLIP to Prefer Descriptions Over Captions (Zur et al., EMNLP 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.emnlp-main.1125.pdf