Research map / Vision-language and multimodal / Cross-modal learning
CODER: Coupled Diversity-Sensitive Momentum Contrastive Learning for Image-Text Retrieval
Haoran Wang, Dongliang He, Wenhao Wu, Boyang Xia, Min Yang, Fu Li, Yunlong Yu, Zhong Ji, Errui Ding, Jingdong Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace BTR loss with the diversity-sensitive contrastive (DCL) loss (ablation) | CODER baseline · Flickr30K | R@1: 78.7 -> 81.9 text, 58.6 -> 61.5 image | Improved |
| add coupled memory banks (M-DCL) to the DCL loss (ablation) | CODER · Flickr30K | R@1: 81.9 -> 82.6 text, 61.5 -> 62.6 image | Improved |
| add concept-level DCL loss (ablation) | CODER · Flickr30K | R@1: 82.6 -> 82.8 text, 62.6 -> 62.8 image | Improved |
| add pseudo-clustering label prediction (PGC) loss (ablation) | CODER · Flickr30K | R@1: 82.8 -> 83.2 text, 62.8 -> 63.1 image | Improved |
| add explicit diversity estimation to implicit DCL (ablation) | DCL loss · Flickr30K | R@1: 80.3 -> 81.5 text, 60.2 -> 61.2 image | Improved |
| use memory bank items in explicit diversity estimation (ablation) | DCL loss · Flickr30K | R@1: 81.5 -> 81.9 text, 61.2 -> 61.5 image | Improved |
| combine M-DCL with instance-level DCL loss (ablation) | CODER · Flickr30K | R@1: 81.9 -> 83.2 text, 61.5 -> 63.1 image | Improved |
| add M-DCL when batch size shrinks from 128 to 32 (ablation) | CODER · Flickr30K | image/text R@1: without M-DCL 61.5 -> 59.3 image, 81.9 -> 79.3 text; with M-DCL only 0.9% drop | Improved |
| vary cluster number K in PGC loss from 5000 to 20000 (ablation) | CODER · Flickr30K | R@1: text R@1 82.9, 83.2, 83.2, 82.8 for K 5000 to 20000 | No clear effect |
| CODER versus best competitor SGRAF | CODER · Flickr30K | R@sum: 499.6 -> 520.9 | Improved |
| CODER versus best competitor SGRAF | CODER · MSCOCO 1K | R@1 text retrieval: 79.6 -> 82.1 | Improved |
Closest papers by meaning
- Image2Sentence based Asymmetrical Zero-shot Composed Image Retrieval (ICLR 2024 · similarity 0.87)
- Retrieval-Enhanced Contrastive Vision-Text Models (ICLR 2024 · similarity 0.86)
- Language-only Training of Zero-shot Composed Image Retrieval (CVPR 2024 · similarity 0.86)
- ConText-CIR: Learning from Concepts in Text for Composed Image Retrieval (CVPR 2025 · similarity 0.86)
- TSVC: Tripartite Learning with Semantic Variation Consistency for Robust Image-Text Retrieval (AAAI 2025 · similarity 0.86)
- CoLLM: A Large Language Model for Composed Image Retrieval (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
