Research map / Optimization / Optimizers
Deep Metric Learning to Rank
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 FastAP instead of the strongest reported baseline | Deep metric learning · Stanford Online Products | R@1: 76.3 -> 76.4 | Helped |
| Use FastAP instead of the strongest reported baseline | Deep metric learning · In-Shop Clothes Retrieval | R@1: 87.3 -> 90.9 | Helped |
| Use FastAP instead of the strongest reported baseline | Deep metric learning · PKU VehicleID large | R@1: 83.1 -> 87.5 | Helped |
| Increase the training batch size (ablation) | FastAP with ResNet-18 · Stanford Online Products, In-Shop Clothes Retrieval, and PKU VehicleID large | R@1 | Helped |
| Use distance quantization with around 10 histogram bins rather than finer quantization (ablation) | FastAP with ResNet-18 · Stanford Online Products, In-Shop Clothes Retrieval, and PKU VehicleID large | R@1 | Helped |
| Increase histogram bin count beyond the peak-performing setting (ablation) | FastAP with ResNet-18 · Stanford Online Products, In-Shop Clothes Retrieval, and PKU VehicleID large | R@1 | Hurt |
| Use only one histogram bin (ablation) | FastAP · Image retrieval | R@1 | Hurt |
| Sample classes from a small number of related categories rather than sampling classes purely at random (ablation) | FastAP minibatch sampling · Few-shot image retrieval | R@1 | Helped |
Closest papers by meaning
- Optimizing Rank-Based Metrics With Blackbox Differentiation (CVPR 2020 · similarity 0.88)
- DarkRank: Accelerating Deep Metric Learning via Cross Sample Similarities Transfer (AAAI 2018 · similarity 0.87)
- Ranked List Loss for Deep Metric Learning (CVPR 2019 · similarity 0.87)
- Deep Metric Learning via Adaptive Learnable Assessment (CVPR 2020 · similarity 0.86)
- Revisiting Training Strategies and Generalization Performance in Deep Metric Learning (ICML 2020 · similarity 0.86)
- Towards Improved Proxy-Based Deep Metric Learning via Data-Augmented Domain Adaptation (AAAI 2024 · 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.