Research map / Label-efficient and robust learning / Recognition and long-tail
Ensemble Deep Manifold Similarity Learning Using Hard Proxies
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 EDMS with Random Walk, optimized hard proxies, and contextual manifold loss | Deep metric learning · CUB-200-2011 | R@1: 63.9 -> 66.1 | Helped |
| Use EDMS with Random Walk, optimized hard proxies, and contextual manifold loss | Deep metric learning · Cars196 | R@1: 86.0 -> 87.6 | Helped |
| Use EDMS with Random Walk, optimized hard proxies, and contextual manifold loss | Deep metric learning · Stanford Online Products | R@1: 76.3 -> 78.5 | Helped |
| Add proxies to Random Walk manifold learning (ablation) | EDMS on CUB-200-2011 · CUB-200-2011 | R@1: 56.1 -> 60.3 | Helped |
| Add optimized hard proxies instead of initial random proxies (ablation) | EDMS with Random Walk on CUB-200-2011 · CUB-200-2011 | R@1: 60.3 -> 66.1 | Helped |
| Add Random Walk manifold similarities (ablation) | EDMS with hard proxies on CUB-200-2011 · CUB-200-2011 | R@1: 63.7 -> 66.1 | Helped |
| Use contextual rather than intrinsic manifold loss (ablation) | EDMS with Random Walk and hard proxies on CUB-200-2011 · CUB-200-2011 | R@1: 61.1 -> 66.1 | Helped |
| Use manifold similarities without proxies (ablation) | EDMS on CUB-200-2011 · CUB-200-2011 | R@1: 58.7 -> 56.1 | Hurt |
| Add Random Walk manifold similarities to proxy loss (ablation) | EDMS with hard proxies on CUB-200-2011 · CUB-200-2011 | R@1: 63.7 -> 66.1 | Helped |
Closest papers by meaning
- Deep Networks and the Multiple Manifold Problem (ICLR 2021 · similarity 0.87)
- Fast Spectral Ranking for Similarity Search (CVPR 2018 · similarity 0.87)
- Guided Similarity Separation for Image Retrieval (NeurIPS 2019 · similarity 0.86)
- Stabilizing Sample Similarity in Representation via Mitigating Random Consistency (ICML 2025 · similarity 0.86)
- Manifold Random Features (NeurIPS 2026 · similarity 0.85)
- Geometry-Aware Similarity Metrics for Neural Representations on Riemannian and Statistical Manifolds (NeurIPS 2026 · 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.