Conditioned Masked Language and Image Modeling for Image-Text Dense Retrieval
Ziyang Luo, Yadong Xi, Rongsheng Zhang, GongZheng Li, Zeng Zhao, Jing Ma
Abstract
Image-text retrieval is a fundamental cross-modal task that takes image/text as a query to retrieve relevant data of another type. The large-scale two-stream pre-trained models like CLIP have achieved tremendous success in this area. They embed the images and texts into instance representations with two separate encoders, aligning them on the instance-level with contrastive learning. Beyond this, the following works adopt the fine-grained token-level interaction (Masked Language and Image Modeling) to boost performance further. However, the vanilla token-level objectives are not designed to aggregate the image-text alignment information into the instance representations, but the token representations, causing a gap between pre-training and application. To address this issue, we carefully design two novel conditioned token-level pre-training objectives, Conditioned Masked Language and Image Modeling (ConMLM and ConMIM), forcing models to aggregate the token-level alignment information into the instance representations. Combing with the instance-level contrastive learning, we propose our cross-modal dense retrieval framework, Conditioned Language-Image Pre-training (ConLIP). Experimental results on two popular cross-modal retrieval benchmarks (MSCOCO and Flickr30k) reveal the effectiveness of our methods.- Anthology ID:
- 2022.findings-emnlp.10
- Volume:
- Findings of the Association for Computational Linguistics: EMNLP 2022
- Month:
- December
- Year:
- 2022
- Address:
- Abu Dhabi, United Arab Emirates
- Venue:
- Findings
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 130–140
- Language:
- URL:
- https://aclanthology.org/2022.findings-emnlp.10
- DOI:
- Cite (ACL):
- Ziyang Luo, Yadong Xi, Rongsheng Zhang, GongZheng Li, Zeng Zhao, and Jing Ma. 2022. Conditioned Masked Language and Image Modeling for Image-Text Dense Retrieval. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 130–140, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
- Cite (Informal):
- Conditioned Masked Language and Image Modeling for Image-Text Dense Retrieval (Luo et al., Findings 2022)
- PDF:
- https://preview.aclanthology.org/ingestion-script-update/2022.findings-emnlp.10.pdf