Research map / Detection and segmentation / CNN features
SEMICON: A Learning-to-Hash Solution for Large-Scale Fine-Grained Image Retrieval
Yang Shen, Xuhao Sun, Xiu-Shen Wei, Qing-Yuan Jiang, Jian Yang
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 SEMICON- (attention maps and global/local hashing branches) to vanilla backbone (ablation) | ResNet-50 hashing · Aircraft (24 bits) | mAP (%): 23.09 -> 70.14 | Improved |
| add SEMICON- to vanilla backbone (ablation) | ResNet-50 hashing · CUB200-2011 (12 bits) | mAP (%): 20.03 -> 34.93 | Improved |
| add SEMICON- to vanilla backbone (ablation) | ResNet-50 hashing · Food101 (24 bits) | mAP (%): 40.93 -> 72.75 | Improved |
| add SEM attention (ablation) | SEMICON- · CUB200-2011 (12 bits) | mAP (%): 34.93 -> 36.58 | Improved |
| add SEM attention (ablation) | SEMICON- · CUB200-2011 (24 bits) | mAP (%): 58.73 -> 64.19 | Improved |
| add SEM attention (ablation) | SEMICON- · Aircraft (12 bits) | mAP (%): 34.18 -> 43.36 | Improved |
| add SEM attention (ablation) | SEMICON- · Food101 (24 bits) | mAP (%): 72.75 -> 75.44 | Improved |
| add ICON (ablation) | SEMICON with SEM · Aircraft (12 bits) | mAP (%): 43.36 -> 49.87 | Improved |
| add ICON (ablation) | SEMICON with SEM · Food101 (12 bits) | mAP (%): 44.95 -> 50.00 | Improved |
| add ICON (ablation) | SEMICON with SEM · CUB200-2011 (24 bits) | mAP (%): 64.19 -> 65.41 | Improved |
| add ICON (ablation) | SEMICON with SEM · Aircraft (32 bits) | mAP (%): 80.64 -> 80.45 | Hurt |
| add ICON (ablation) | SEMICON with SEM · Food101 (48 bits) | mAP (%): 82.40 -> 82.44 | No clear effect |
Closest papers by meaning
- A$^2$-Net: Learning Attribute-Aware Hash Codes for Large-Scale Fine-Grained Image Retrieval (NeurIPS 2021 · similarity 0.89)
- ExchNet: A Unified Hashing Network for Large-Scale Fine-Grained Image Retrieval (ECCV 2020 · similarity 0.89)
- Learning Attribute-Aware Hash Codes for Fine-Grained Image Retrieval via Query Optimization (ICML 2025 · similarity 0.88)
- Shared Latent Characteristic Anchored Hash Codes Generation for Efficient Fine-Grained Image Retrieval (TPAMI 2026 · similarity 0.88)
- Characteristics Matching Based Hash Codes Generation for Efficient Fine-grained Image Retrieval (CVPR 2024 · similarity 0.87)
- An Asymmetric Augmented Self-Supervised Learning Method for Unsupervised Fine-Grained Image Hashing (CVPR 2024 · 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.
