Research map / Architectures and efficiency / CNN design
Efficient Large-Scale Similarity Search Using Matrix Factorization
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 sparse dictionary-learning decoding instead of eigendecomposition | The proposed group-testing search · Oxford105k | mAP: 8.0 -> 65.5 | Helped |
| Use sparse dictionary-learning decoding instead of eigendecomposition | The proposed group-testing search · Paris106k | mAP: 40.9 -> 78.9 | Helped |
| Use eigendecomposition instead of dictionary learning | The proposed group-testing search · Holidays | mAP: 55.2 -> 76.9 | Helped |
| Use dictionary learning with sparse decoding | Image retrieval using R-MAC features · Oxford5k | mAP: 66.9 -> 73.7 | Helped |
| Use dictionary learning with sparse decoding | Image retrieval using R-MAC features · Paris106k | mAP: 75.7 -> 78.9 | Helped |
| Use dictionary learning with sparse decoding | Image retrieval using R-MAC features · Oxford105k | mAP: 61.6 -> 65.5 | Helped |
| Increase dictionary size from M = N/200 to M = N/50 at m = 100 (ablation) | Dictionary-learning retrieval · Yahoo100M | mAP: 85.7 -> 92.8 | Helped |
| Increase sparsity level from m = 20 to m = 100 at M = N/100 (ablation) | Dictionary-learning retrieval · Yahoo100M | mAP: 71.4 -> 89.4 | Helped |
| Train the dictionary on the full dataset rather than a coreset of size N/10 (ablation) | Dictionary-learning retrieval · Oxford105k | mAP: 60.1±1.1 -> 65.5 | Helped |
| Train the dictionary on the full dataset rather than a coreset of size N/10 (ablation) | Dictionary-learning retrieval · Paris106k | mAP: 78.3±1.0 -> 81.2 | Helped |
| Increase sparsity level m (ablation) | Dictionary-learning retrieval · Oxford105k | mAP | Helped |
| Vary sparsity level m (ablation) | Dictionary-learning retrieval · Most evaluated datasets and dictionary sizes | mAP | No clear effect |
Closest papers by meaning
- Guided Similarity Separation for Image Retrieval (NeurIPS 2019 · similarity 0.88)
- Link and Code: Fast Indexing With Graphs and Compact Regression Codes (CVPR 2018 · similarity 0.86)
- Explore-Exploit Graph Traversal for Image Retrieval (CVPR 2019 · similarity 0.86)
- Fast Spectral Ranking for Similarity Search (CVPR 2018 · similarity 0.86)
- Exploiting Distribution Constraints for Scalable and Efficient Image Retrieval (ICLR 2025 · similarity 0.85)
- Instance-Level Composed Image Retrieval (NeurIPS 2025 · similarity 0.84)
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.