Research map / Label-efficient and robust learning / Recognition and long-tail
Signal-To-Noise Ratio: A Robust Distance Metric for Deep Metric Learning
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 |
|---|---|---|---|
| Replace Euclidean distance with SNR distance in triplet loss | Deep metric learning with AlexNet · CARS196, embedding size 64 | Recall@1: 40.6 -> 49.1 | Helped |
| Replace Euclidean distance with SNR distance in lifted loss | Deep metric learning with AlexNet · CARS196, embedding size 64 | Recall@1: 40.6 -> 43.8 | Helped |
| Replace Euclidean distance with SNR distance in contrastive loss | Deep metric learning with AlexNet · CUB200-2011, embedding size 64 | Recall@1: 29.7 -> 33.1 | Helped |
| Replace Euclidean distance with SNR distance in triplet loss | Deep metric learning with AlexNet · CUB200-2011, embedding size 64 | Recall@1: 32.6 -> 51.6 | Helped |
| Replace Euclidean distance with SNR distance in N-pair loss | Deep metric learning with AlexNet · CARS196, embedding size 64 | Recall@1: 38.3 -> 44.9 | Helped |
| Replace Euclidean distance with SNR distance in N-pair loss | Deep metric learning with AlexNet · CUB200-2011, embedding size 64 | Recall@1: 43.1 -> 48.5 | Helped |
| Use SNR distance rather than Euclidean distance | Contrastive metric learning · CIFAR10, embedding size 64, Euclidean ranking | MAP@59000: 69.3 -> 79.0 | Helped |
| Use SNR-based contrastive loss for hashing | DSNRH with CNN-F · CIFAR10, 16 bits | MAP@50000: 0.915 -> 0.925 | Helped |
| Use SNR-based contrastive loss for hashing | DSNRH with CNN-F · NUS-WIDE, 16 bits | MAP@50000: 0.756 -> 0.830 | Helped |
Closest papers by meaning
- Divide and Conquer the Embedding Space for Metric Learning (CVPR 2019 · similarity 0.89)
- Deep Metric Learning via Facility Location (CVPR 2017 · similarity 0.88)
- Denoising Multi-Similarity Formulation: A Self-Paced Curriculum-Driven Approach for Robust Metric Learning (AAAI 2023 · similarity 0.88)
- Local Similarity-Aware Deep Feature Embedding (NeurIPS 2016 · similarity 0.88)
- Energy Confused Adversarial Metric Learning for Zero-Shot Image Retrieval and Clustering (AAAI 2019 · similarity 0.88)
- Deep Compositional Metric Learning (CVPR 2021 · 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.