Research map / Video / Video segmentation and tracking
Central Similarity Quantization for Efficient Image and Video Retrieval
Check 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 central similarity with hash centers instead of pairwise similarity (ablation) | Deep supervised hashing · ImageNet | mAP: 0.655 -> 0.873 | Helped |
| Use central similarity with hash centers instead of pairwise similarity (ablation) | Deep video hashing · UCF101 | mAP: 0.716 -> 0.838 | Helped |
| Add pairwise similarity loss L_P to CSQ (ablation) | CSQ · ImageNet | mAP: 0.851 -> 0.847; 0.865 -> 0.870; 0.873 -> 0.871 | Mixed |
| Add pairwise similarity loss L_P to CSQ (ablation) | CSQ · UCF101 | mAP: 0.838 -> 0.840; 0.875 -> 0.868; 0.874 -> 0.881 | Mixed |
| Remove quantization loss L_Q (ablation) | CSQ · UCF101 | mAP: 0.875 -> 0.854 | Hurt |
| Use precomputed hash centers rather than learned feature-based centers (ablation) | CSQ · ImageNet | mAP: 0.827 -> 0.873 | Helped |
| Use precomputed hash centers rather than learned feature-based centers (ablation) | CSQ · UCF101 | mAP: 0.835 -> 0.874 | Helped |
| Apply CSQ rather than DCH | Image retrieval · ImageNet | mAP: 0.652 -> 0.851; 0.737 -> 0.865; 0.758 -> 0.873 | Helped |
| Apply CSQ rather than DCH | Image retrieval · MS COCO | mAP: 0.759 -> 0.796; 0.801 -> 0.838; 0.825 -> 0.861 | Helped |
| Apply CSQ rather than DCH | Image retrieval · NUS-WIDE | mAP: 0.773 -> 0.810; 0.795 -> 0.825; 0.818 -> 0.839 | Helped |
| Apply CSQ rather than SRH | Video retrieval · UCF101 | mAP: 0.716 -> 0.838; 0.692 -> 0.875; 0.754 -> 0.874 | Helped |
| Apply CSQ rather than SRH | Video retrieval · HMDB51 | mAP: 0.491 -> 0.527; 0.503 -> 0.565; 0.509 -> 0.579 | Helped |
Closest papers by meaning
- Deep Hashing With Minimal-Distance-Separated Hash Centers (CVPR 2023 · similarity 0.94)
- SADIH: Semantic-Aware DIscrete Hashing (AAAI 2019 · similarity 0.90)
- Statistical Model-driven Similarity Hashing: Bridging Modalities for Efficient Unsupervised Retrieval (AAAI 2025 · similarity 0.87)
- Vision-guided Text Mining for Unsupervised Cross-modal Hashing with Community Similarity Quantization (AAAI 2025 · similarity 0.87)
- Effective Comparative Prototype Hashing for Unsupervised Domain Adaptation (AAAI 2024 · similarity 0.86)
- Pairwise-Label-Based Deep Incremental Hashing with Simultaneous Code Expansion (AAAI 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.