Research map / Continual learning / Online and replay-based
Asynchronous Teacher Guided Bit-wise Hard Mining for Online Hashing
Sheng Jin, Qin Zhou, Hongxun Yao, Yao Liu, Xian-Sheng Hua
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 ATHOH instead of HCOH | Online hashing · CIFAR-10, 32 bits | MAP: 0.698 -> 0.734 | Improved |
| Use ATHOH instead of HCOH | Online hashing · NUS-WIDE, 32 bits | MAP: 0.692 -> 0.717 | Improved |
| Use ATHOH instead of HCOH | Online hashing · CIFAR-10, 32 bits | Precision@top500: 0.756 -> 0.791 | Improved |
| Use ATHOH instead of HCOH | Online hashing · NUS-WIDE, 32 bits | Precision@top500: 0.794 -> 0.825 | Improved |
| Replace singular-vector codewords with Gaussian-sampled codewords (ablation) | ATHOH · CIFAR-10, 32 bits | MAP: 0.702 -> 0.692 | Hurt |
| Replace singular-vector codewords with Bernoulli-sampled codewords (ablation) | ATHOH · CIFAR-10, 32 bits | MAP: 0.702 -> 0.695 | Hurt |
| Replace singular-vector codewords with Hadamard codewords (ablation) | ATHOH · CIFAR-10, 32 bits | MAP: 0.702 -> 0.704 | No clear effect |
| Add bit-wise attention (ablation) | ATHOH · CIFAR-10, 32 bits | MAP: 0.702 -> 0.728 | Improved |
| Add bit-wise attention (ablation) | ATHOH · NUS-WIDE, 32 bits | MAP: 0.690 -> 0.708 | Improved |
| Add global knowledge distillation (ablation) | ATHOH · CIFAR-10, 32 bits | MAP: 0.702 -> 0.725 | Improved |
| Update the teacher asynchronously rather than synchronously (ablation) | ATHOH teacher-student framework · CIFAR-10, 32 bits | MAP: 0.712 -> 0.725 | Improved |
| Use a long-term asynchronous teacher instead of a short-term teacher (ablation) | ATHOH teacher-student framework · CIFAR-10, 32 bits | MAP: 0.720 -> 0.725 | Improved |
Closest papers by meaning
- Fast Class-wise Updating for Online Hashing (TPAMI 2020 · similarity 0.88)
- MIHash: Online Hashing With Mutual Information (ICCV 2017 · similarity 0.87)
- Online Hashing with Efficient Updating of Binary Codes (AAAI 2020 · similarity 0.87)
- Towards Optimal Discrete Online Hashing with Balanced Similarity (AAAI 2019 · similarity 0.86)
- KALAHash: Knowledge-Anchored Low-Resource Adaptation for Deep Hashing (AAAI 2025 · similarity 0.84)
- Pairwise-Label-Based Deep Incremental Hashing with Simultaneous Code Expansion (AAAI 2024 · similarity 0.83)
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.
