Research map / Label-efficient and robust learning / Recognition and long-tail
Fine-Grained Image Retrieval via Dual-Vision Adaptation
Xin Jiang, Meiqi Cao, Hao Tang, Fei Shen, Zechao Li
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 In-Context Adaptation to frozen ViT (ablation) | DVA · CUB-200-2011 | Recall@1: 72.5 -> 83.6 | Improved |
| add Object-Perceptual Adaptation at inference (ablation) | DVA (Base+ICA) · CUB-200-2011 | Recall@1: 83.6 -> 86.8 | Improved |
| add Discriminative Perceptual Transfer instead of OPA at inference (ablation) | DVA · CUB-200-2011 | Recall@1: 86.8 -> 84.9 | Mixed |
| add object perception (background category) to discriminative perception (ablation) | DVA · CUB-200-2011 | Recall@1: 84.1 -> 84.9 | Improved |
| tune attention projector only vs whole ViT (ablation) | ViT · CUB-200-2011 | Recall@1: 82.7 -> 82.7 | No clear effect |
| use IC on both Q and K instead of Q alone (ablation) | ICA · CUB-200-2011 | Recall@1: 82.7 -> 84.9 | Improved |
| add IC on V to Q,K (ablation) | ICA · CUB-200-2011 | Recall@1: 84.9 -> 84.4 | Hurt |
| increase beta above 3 (ablation) | DVA · CUB-200-2011 | Recall@1 | Hurt |
| DVA vs Hyp-ViT | FGIR · CUB-200-2011 closed-set | Recall@1: 84.2 -> 85.2 | Improved |
| DVA vs HIST on Stanford Cars | FGIR · Stanford Cars closed-set | Recall@1: 89.2 -> 88.0 | Hurt |
Closest papers by meaning
- Fine-Grained Retrieval Prompt Tuning (AAAI 2023 · similarity 0.91)
- Fine-Grained Recognition in the Wild: A Multi-Task Domain Adaptation Approach (ICCV 2017 · similarity 0.89)
- Language-driven Fine-grained Retrieval (CVPR 2026 · similarity 0.89)
- Analyzing and Boosting the Power of Fine-Grained Visual Recognition for Multi-modal Large Language Models (ICLR 2025 · similarity 0.86)
- Learning to Align Generative Appearance Priors for Fine-grained Image Retrieval (NeurIPS 2026 · similarity 0.86)
- DiVE-k: DIFFERENTIAL VISUAL REASONING FOR FINE-GRAINED IMAGE RECOGNITION (ICLR 2026 · 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.
