Research map / Optimization / Optimizers
Learning to Hash with Binary Deep Neural Network
Thanh-Toan Do, Anh-Dzung Doan, Ngai-Man Cheung
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 |
|---|---|---|---|
| use UH-BDNN over the binary autoencoder BA on CIFAR10 | UH-BDNN · CIFAR10 (unsupervised, 32 bits) | precision@2 (Hamming radius 2, %): 17.00 -> 18.35 | Improved |
| use UH-BDNN over BA on MNIST | UH-BDNN · MNIST (unsupervised, 32 bits) | precision@2 (Hamming radius 2, %): 35.29 -> 38.50 | Improved |
| use UH-BDNN over BA on SIFT1M | UH-BDNN · SIFT1M (unsupervised, 32 bits) | precision@2 (Hamming radius 2, %): 77.15 -> 80.55 | Improved |
| use UH-BDNN over BA at low code length | UH-BDNN · CIFAR10 (unsupervised, 8 bits) | precision@2 (Hamming radius 2, %): 0.55 -> 0.55 | No clear effect |
| use UH-BDNN over Deep Hashing (DH) | UH-BDNN · CIFAR10 (DH protocol, 16 bits) | mAP (%): 16.17 -> 17.83 | Improved |
| use UH-BDNN over Deep Hashing (DH) | UH-BDNN · MNIST (DH protocol, 16 bits) | mAP (%): 43.14 -> 45.38 | Improved |
| use SH-BDNN over SDH at 8 bits | SH-BDNN · CIFAR10 (supervised, 8 bits) | precision@2 (Hamming radius 2, %): 31.60 -> 54.12 | Improved |
| use SH-BDNN over SDH at 8 bits | SH-BDNN · MNIST (supervised, 8 bits) | precision@2 (Hamming radius 2, %): 36.49 -> 84.26 | Improved |
| use SH-BDNN over SDH at higher code length | SH-BDNN · MNIST (supervised, 32 bits) | precision@2 (Hamming radius 2, %): 94.43 -> 95.51 | No clear effect |
| use SH-BDNN over DRSCH, a CNN-based hashing method | SH-BDNN · CIFAR10 (CNN-hashing protocol, 16 bits) | mAP (%): 61.46 -> 64.30 | Improved |
| use SH-BDNN over DSRH, a CNN-based hashing method | SH-BDNN · CIFAR10 (CNN-hashing protocol, 16 bits) | mAP (%): 60.84 -> 64.30 | Improved |
Closest papers by meaning
- Unsupervised Deep Hashing with Similarity-Adaptive and Discrete Optimization (TPAMI 2018 · similarity 0.91)
- From Hashing to CNNs: Training Binary Weight Networks via Hashing (AAAI 2018 · similarity 0.91)
- One Loss for All: Deep Hashing with a Single Cosine Similarity based Learning Objective (NeurIPS 2021 · similarity 0.90)
- Greedy Hash: Towards Fast Optimization for Accurate Hash Coding in CNN (NeurIPS 2018 · similarity 0.90)
- Deep Hashing via Discrepancy Minimization (CVPR 2018 · similarity 0.90)
- Deep Supervised Discrete Hashing (NeurIPS 2017 · similarity 0.89)
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.
