Research map / Learning theory and algorithms / Matrix and tensor methods
Data-Adaptive Metric Learning with Scale Alignment
Shuo Chen, Chen Gong, Jian Yang, Ying Tai, Le Hui, Jun Li
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 data-adaptive projection (DAML) instead of a single fixed projection matrix (ITML) | DAML · UCI classification (MNIST) | classification error (%): 14.31 -> 11.34 | Improved |
| Use data-adaptive projection (DAML) instead of a single fixed projection matrix (ITML) | DAML · Letters | classification error (%): 6.24 -> 3.11 | Improved |
| Use data-adaptive projection (DAML) instead of a single fixed projection matrix (ITML) | DAML · Autompg | classification error (%): 26.62 -> 23.95 | Improved |
| Use data-adaptive projection (DAML) instead of a single fixed projection matrix (ITML) | DAML · IsoLet | classification error (%): 9.23 -> 2.98 | Improved |
| Kernelize DAML (KDAML) with a Gaussian kernel | DAML · Autompg | classification error (%): 23.95 -> 21.45 | Improved |
| Kernelize DAML (KDAML) with a Gaussian kernel | DAML · MNIST | classification error (%): 11.34 -> 11.12 | Improved |
| Kernelize DAML (KDAML) with a Gaussian kernel | DAML · Australia | classification error (%): 13.17 -> 13.11 | Improved |
| Kernelize DAML (KDAML) with a Gaussian kernel | DAML · PubFig retrieval | top-5 classification error (%): 12.95 -> 12.89 | Improved |
| Kernelize DAML (KDAML) with a Gaussian kernel | DAML · OSR retrieval | top-5 classification error (%): 23.41 -> 23.18 | Improved |
| Kernelize DAML (KDAML) with a Gaussian kernel | DAML · PubFig verification | AUC: 0.90746 -> 0.9242 | Improved |
| Kernelize DAML (KDAML) with a Gaussian kernel | DAML · MVS verification | AUC: 0.75373 -> 0.76998 | Improved |
| Use DAML instead of the best retrieval baseline LGMML | DAML · PubFig retrieval | top-5 classification error (%): 13.15 -> 12.95 | Improved |
Closest papers by meaning
- Deep Metric Learning with Adaptively Composite Dynamic Constraints (TPAMI 2023 · similarity 0.86)
- Denoising Multi-Similarity Formulation: A Self-Paced Curriculum-Driven Approach for Robust Metric Learning (AAAI 2023 · similarity 0.86)
- Robust Metric Learning on Grassmann Manifolds with Generalization Guarantees (AAAI 2019 · similarity 0.86)
- Orthogonality-Promoting Distance Metric Learning: Convex Relaxation and Theoretical Analysis (ICML 2018 · similarity 0.86)
- Learning Low-Dimensional Metrics (NeurIPS 2017 · similarity 0.86)
- Variational Metric Scaling for Metric-Based Meta-Learning (AAAI 2020 · similarity 0.85)
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.
