Research map / Label-efficient and robust learning / Recognition and long-tail
Multi-level Distance Regularization for Deep Metric Learning
Read the paperCheck 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 |
|---|---|---|---|
| Add MDR to Triplet loss | IBN with Triplet loss · CUB-200 | Recall@1: 65.1±0.3 -> 68.8±0.5 | Helped |
| Add MDR to Triplet loss | IBN with Triplet loss · Cars-196 | Recall@1: 79.8±0.3 -> 88.5±0.3 | Helped |
| Add MDR to Triplet loss | IBN with Triplet loss · SOP | Recall@1: 79.1±0.1 -> 80.1±0.0 | Helped |
| Add MDR to Triplet loss | IBN with Triplet loss · In-Shop | Recall@1: 90.1±0.1 -> 90.5±0.1 | Helped |
| Add MDR (ablation) | ResNet18 with Triplet loss · CUB-200 | Recall@1: 51.2±0.5 -> 63.6±0.5 | Helped |
| Add MDR (ablation) | ResNet50 with Triplet loss · CUB-200 | Recall@1: 58.5±0.4 -> 65.8±0.3 | Helped |
| Add MDR (ablation) | ResNet18 with Triplet loss · Cars-196 | Recall@1: 63.4±1.0 -> 82.3±0.2 | Helped |
| Add MDR (ablation) | ResNet50 with Triplet loss · Cars-196 | Recall@1: 77.5±0.4 -> 87.6±0.2 | Helped |
| Add MDR to Contrastive loss (ablation) | IBN · CUB-200 | Recall@1: 63.9±0.3 -> 65.6±0.2 | Helped |
| Add MDR to Contrastive loss (ablation) | IBN · Cars-196 | Recall@1: 83.2±0.1 -> 86.1±1.0 | Helped |
| Add MDR to Margin loss (ablation) | IBN · CUB-200 | Recall@1: 59.3±0.5 -> 67.5±0.3 | Helped |
| Add MDR to Margin loss (ablation) | IBN · Cars-196 | Recall@1: 79.1±0.3 -> 88.2±0.4 | Helped |
Closest papers by meaning
- Moving in the Right Direction: A Regularization for Deep Metric Learning (CVPR 2020 · similarity 0.88)
- Orthogonality-Promoting Distance Metric Learning: Convex Relaxation and Theoretical Analysis (ICML 2018 · similarity 0.87)
- Multi-Similarity Loss With General Pair Weighting for Deep Metric Learning (CVPR 2019 · similarity 0.87)
- Non-Isotropy Regularization for Proxy-Based Deep Metric Learning (CVPR 2022 · similarity 0.87)
- Divide and Conquer the Embedding Space for Metric Learning (CVPR 2019 · similarity 0.87)
- Improved Deep Metric Learning with Multi-class N-pair Loss Objective (NeurIPS 2016 · similarity 0.87)
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.