Research map / Label-efficient and robust learning / Recognition and long-tail
Deep Hash Distillation for Image Retrieval
Young Kyun Jang, Geonmo Gu, Byungsoo Ko, Isaac Kang, Nam Ik Cho
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 self-distilled hashing (ablation) | HashNet (AlexNet) · ImageNet 16-bit | mAP: 0.337 -> 0.501 | Improved |
| add self-distilled hashing (ablation) | DCH (AlexNet) · ImageNet 16-bit | mAP: 0.571 -> 0.640 | Improved |
| add self-distilled hashing (ablation) | CSQ (AlexNet) · ImageNet 16-bit | mAP: 0.569 -> 0.634 | Improved |
| add self-distilled hashing (ablation) | DPN (AlexNet) · NUS-WIDE 16-bit | mAP: 0.753 -> 0.757 | No clear effect |
| add self-distilled hashing (ablation) | hash proxy loss (AlexNet) · ImageNet 16-bit | mAP: 0.574 -> 0.642 | Improved |
| add self-distilled hashing (ablation) | hash proxy loss (AlexNet) · MS COCO 16-bit | mAP: 0.706 -> 0.725 | Improved |
| add BCE-based quantization loss to hash proxy loss (ablation) | DHD without SdH (AlexNet) · ImageNet 16-bit | mAP: 0.574 -> 0.583 | Improved |
| add BCE-based quantization loss to hash proxy loss (ablation) | DHD with SdH (AlexNet) · ImageNet 16-bit | mAP: 0.642 -> 0.657 | Improved |
| add self-distilled hashing (ablation) | DHD (ResNet50) · ImageNet, no deformation | mAP: 0.871 -> 0.891 | Improved |
| add self-distilled hashing (ablation) | DHD (ResNet50) · ImageNet unseen zoom in | mAP: 0.552 -> 0.658 | Improved |
| add self-distilled hashing (ablation) | DHD (ResNet50) · ImageNet unseen Gaussian noise | mAP: 0.673 -> 0.768 | Improved |
| add self-distilled hashing (ablation) | DHD (ResNet50) · ImageNet unseen zoom out | mAP: 0.805 -> 0.816 | Improved |
Closest papers by meaning
- Deep Hashing Network for Efficient Similarity Retrieval (AAAI 2016 · similarity 0.93)
- Deep Supervised Hashing for Fast Image Retrieval (CVPR 2016 · similarity 0.92)
- Deep Quantization Network for Efficient Image Retrieval (AAAI 2016 · similarity 0.92)
- Deep Incremental Hashing Network for Efficient Image Retrieval (CVPR 2019 · similarity 0.91)
- One Loss for Quantization: Deep Hashing With Discrete Wasserstein Distributional Matching (CVPR 2022 · similarity 0.90)
- Pairwise-Label-Based Deep Incremental Hashing with Simultaneous Code Expansion (AAAI 2024 · similarity 0.90)
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.
