Research map / Label-efficient and robust learning / Recognition and long-tail
Comparator Networks
Weidi Xie, Li Shen, Andrew Zisserman
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 |
|---|---|---|---|
| add Deep Comparator Network (diversity regularizer) on top of SENet50 | SENet50 · IJB-B | TAR at FAR=1E-4: 0.800 -> 0.849 | Improved |
| add Deep Comparator Network (keypoints regularizer) on top of ResNet50 | ResNet50 · IJB-B | TAR at FAR=1E-4: 0.784 -> 0.850 | Improved |
| add Deep Comparator Network (diversity regularizer) on top of SENet50 | SENet50 · IJB-C | TAR at FAR=1E-4: 0.840 -> 0.885 | Improved |
| add Deep Comparator Network (diversity regularizer) on top of ResNet50 | ResNet50 · IJB-C | TAR at FAR=1E-4: 0.825 -> 0.880 | Improved |
| ensemble ResNet50 with SENet50 | single image classification models · IJB-B | TAR at FAR=1E-4: 0.800 -> 0.800 | No clear effect |
| ensemble ResNet50 with SENet50 | single image classification models · IJB-C | TAR at FAR=1E-4: 0.840 -> 0.841 | No clear effect |
| replace diversity regularizer with keypoints regularizer (ablation) | SENet50+DCN · IJB-B | TAR at FAR=1E-4: 0.849 -> 0.846 | No clear effect |
| replace diversity regularizer with keypoints regularizer (ablation) | ResNet50+DCN · IJB-C | TAR at FAR=1E-4: 0.880 -> 0.867 | Hurt |
| replace diversity regularizer with keypoints regularizer (ablation) | SENet50+DCN · IJB-C | TAR at FAR=1E-1: 0.998 -> 0.998 | No clear effect |
| DCN with hard-sample mining vs multicolumn network MN-vc | SENet50+DCN(Divs) · IJB-B | TAR at FAR=1E-4: 0.831 -> 0.849 | Improved |
Closest papers by meaning
- Quality Aware Network for Set to Set Recognition (CVPR 2017 · similarity 0.85)
- FaceCoresetNet: Differentiable Coresets for Face Set Recognition (AAAI 2024 · similarity 0.84)
- Unconstrained Still/Video-Based Face Verification with Deep Convolutional Neural Networks (IJCV 2017 · similarity 0.83)
- DAM: Discrepancy Alignment Metric for Face Recognition (ICCV 2021 · similarity 0.83)
- Discriminative Covariance Oriented Representation Learning for Face Recognition With Image Sets (CVPR 2017 · similarity 0.83)
- Unlocking Better Closed-Set Alignment Based on Neural Collapse for Open-Set Recognition (AAAI 2025 · similarity 0.81)
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.
