Research map / Vision-language and multimodal / Human and embodied modeling
Dual Compositional Learning in Interactive Image Retrieval
Jongseok Kim, Youngjae Yu, Hoeseong Kim, Gunhee Kim
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 the Correction Network to the Composition Network (ablation) | DCNet · Fashion-IQ validation | Avg: 38.98 -> 40.83 | Improved |
| add the Correction Network to Concat (ablation) | Concat · Fashion-IQ validation | Avg: 32.77 -> 33.54 | Improved |
| add the Correction Network to TIRG (ablation) | TIRG · Fashion-IQ validation | Avg: 33.31 -> 33.87 | Improved |
| add the Correction Network to VAL (ablation) | VAL · Fashion-IQ validation | Avg: 37.13 -> 38.19 | Improved |
| add the joint loss Ljoint (ablation) | DCNet image and text experts · Fashion-IQ validation | Avg: 39.55 -> 39.73 | Improved |
| add the joint loss Ljoint (ablation) | DCNet image and text experts · Fashion-IQ validation | Avg: 39.78 -> 40.31 | Improved |
| combine Composition and Correction Network scores at inference (ablation) | DCNet · Shoes validation | R@10: 53.24 -> 53.82 | Improved |
| combine Composition and Correction Network scores at inference (ablation) | DCNet · Shoes validation | R@50: 78.67 -> 79.33 | Improved |
| use the proposed Fused Difference representation instead of a single text expert (ablation) | Correction Network · Fashion-IQ validation | Avg: 38.75 -> 39.63 | Improved |
| add the Correction Network to TIRG (ablation) | TIRG · Fashion200K | R@10: 45.62 -> 46.31 | Improved |
| add the Correction Network to TIRG (ablation) | TIRG · Fashion200K | R@50: 67.24 -> 67.41 | Improved |
| combine Composition and Correction Network scores at inference (ablation) | DCNet · Fashion-IQ validation | Avg: 40.31 -> 40.83 | Improved |
Closest papers by meaning
- FashionERN: Enhance-and-Refine Network for Composed Fashion Image Retrieval (AAAI 2024 · similarity 0.88)
- Do Composed Image Retrieval Benchmarks Require Multimodal Composition? (NeurIPS 2026 · similarity 0.87)
- CoLLM: A Large Language Model for Composed Image Retrieval (CVPR 2025 · similarity 0.86)
- Vision-by-Language for Training-Free Compositional Image Retrieval (ICLR 2024 · similarity 0.85)
- Decomposing Semantic Shifts for Composed Image Retrieval (AAAI 2024 · similarity 0.85)
- Image2Sentence based Asymmetrical Zero-shot Composed Image Retrieval (ICLR 2024 · 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.
