Research map / Label-efficient and robust learning / Recognition and long-tail
Supervised Deep Hashing for Hierarchical Labeled Data
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use SHDH's hierarchy-weighted similarity and deep hash-code learning rather than DPSH | CIFAR-100 retrieval · CIFAR-100 | ACG@100, DCG@100, and NDCG@100: ACG@100: 0.4643 -> 0.5225 (32 bits), 0.4973 -> 0.5724 (48 bits), 0.5140 -> 0.6084 (64 bits); DCG@100: 11.5129 -> 12.7460 | Helped |
| Use SHDH's hierarchy-weighted similarity and deep hash-code learning rather than DPSH | IAPRTC-12 retrieval · IAPRTC-12 | ACG@100, DCG@100, and NDCG@100: ACG@100: 4.0085 -> 4.4870 (48 bits), 4.0227 -> 4.5284 (64 bits), 4.0980 -> 4.5869 (128 bits); DCG@100: 91.4972 -> 100.63 | Helped |
| Use hierarchical similarity instead of the original similarity labels in COSDISH | COSDISH on CIFAR-100 · CIFAR-100 | ACG@100: 0.1366 -> 0.1345 (32 bits); 0.1428 -> 0.1860 (48 bits); 0.1501 -> 0.2008 (64 bits) | Helped |
| Use hierarchical similarity instead of the original similarity labels in KSH | KSH on CIFAR-100 · CIFAR-100 | ACG@100: 0.1291 -> 0.1611 (32 bits); 0.1393 -> 0.1576 (48 bits); 0.1509 -> 0.1718 (64 bits) | Mixed |
| Use hierarchical similarity instead of the original similarity labels in COSDISH | COSDISH on IAPRTC-12 · IAPRTC-12 | NDCG@100: 0.6412 -> 0.6957 (48 bits); 0.6443 -> 0.6885 (64 bits); 0.6408 -> 0.6970 (128 bits) | Helped |
| Use hierarchical similarity instead of the original similarity labels in KSH | KSH on IAPRTC-12 · IAPRTC-12 | NDCG@100: 0.6507 -> 0.6459 (48 bits); 0.6482 -> 0.6494 (64 bits); 0.6408 -> 0.6518 (128 bits) | Mixed |
| Increase the SHDH code length | SHDH on CIFAR-100 · CIFAR-100 | ACG@100: 0.5225 (32 bits), 0.5724 (48 bits), 0.6084 (64 bits) | Helped |
| Vary alpha within 0.5 to 2 (ablation) | SHDH · CIFAR-100 and IAPRTC-12 | retrieval performance | No clear effect |
Closest papers by meaning
- Deep Hashing Network for Efficient Similarity Retrieval (AAAI 2016 · similarity 0.92)
- Deep Supervised Hashing for Fast Image Retrieval (CVPR 2016 · similarity 0.92)
- Deep Supervised Discrete Hashing (NeurIPS 2017 · similarity 0.91)
- Deep Graph Online Hashing for Multi-Label Image Retrieval (AAAI 2025 · similarity 0.89)
- Generalized Debiased Semi-Supervised Hashing for Large-Scale Image Retrieval (AAAI 2025 · similarity 0.88)
- Pairwise-Label-Based Deep Incremental Hashing with Simultaneous Code Expansion (AAAI 2024 · similarity 0.88)
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.