Research map / Vision-language and multimodal / Cross-modal learning
Toward Robust and Harmonious Adaptation for Cross-modal Retrieval
Haobin Li, Zhuqi Liu, Mouxing Yang, Xi Peng
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 |
|---|---|---|---|
| add query uniformity loss L_U (ablation) | REST · General2COCO | avg Recall@1: 52.4 -> 56.2 | Improved |
| add query-gallery consistency loss L_C (ablation) | REST · General2COCO | avg Recall@1: 52.4 -> 57.2 | Improved |
| add query-gallery gap loss L_G (ablation) | REST · General2COCO | avg Recall@1: 52.4 -> 54.3 | Improved |
| add L_G to L_U + L_C (ablation) | REST · General2COCO | avg Recall@1: 58.0 -> 59.0 | Improved |
| add L_U to L_C + L_G (ablation) | REST · General2COCO | avg Recall@1: 57.8 -> 59.0 | Improved |
| add gradient decoupling to Tent (ablation) | Tent · Fashion-Gen TR@1 | TR@1: 11.7 -> 20.8 | Improved |
| add gradient decoupling to Tent (ablation) | Tent · COCO-D TR@1 | TR@1: 12.8 -> 46.5 | Improved |
| add gradient decoupling to REST (ablation) | REST · Fashion-Gen TR@1 | TR@1: 22.3 -> 23.4 | Improved |
| add gradient decoupling to REST (ablation) | REST · COCO-D TR@1 | TR@1: 50.4 -> 51.8 | Improved |
| add gradient decoupling to REST reduces gradient angle (ablation) | REST · Fashion-Gen | gradient angle (degrees): 98.6 -> 83.2 | Improved |
| add gradient decoupling to Tent reduces gradient angle (ablation) | Tent · Fashion-Gen | gradient angle (degrees): 120.1 -> 86.1 | Improved |
| move from online to diverse query shift (ablation) | Tent · DQS vs OQS benchmarks | avg Recall: 56.1 -> 21.8 | Hurt |
Closest papers by meaning
- Test-time Adaptation for Cross-modal Retrieval with Query Shift (ICLR 2025 · similarity 0.94)
- Learning to Rematch Mismatched Pairs for Robust Cross-Modal Retrieval (CVPR 2024 · similarity 0.86)
- Beyond Global Similarity: Multi-Conditional Retrieval for Fine-Grained Cross-Modal Understanding (CVPR 2026 · similarity 0.86)
- NeighborRetr: Balancing Hub Centrality in Cross-Modal Retrieval (CVPR 2025 · similarity 0.86)
- Adaptive Prompt-Based Semantic Embedding with Inspire Potential of Implicit Knowledge for Cross-Modal Retrieval (AAAI 2025 · similarity 0.85)
- Dynamic Adapter with Semantics Disentangling for Cross-lingual Cross-modal Retrieval (AAAI 2025 · similarity 0.85)
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.
