Research map / Label-efficient and robust learning / Recognition and long-tail
UniformFace: Learning Deep Equidistributed Representation for Face Recognition
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 |
|---|---|---|---|
| Add uniform loss to SphereFace (ablation) | SphereFace · MegaFace Challenge I | Rank-1 identification accuracy (%): 76.65 -> 79.98 | Helped |
| Add uniform loss to SphereFace (ablation) | SphereFace · MegaFace Challenge I | TAR at 10^-6 FAR (%): 92.32 -> 95.36 | Helped |
| Add uniform loss to SphereFace (ablation) | SphereFace · IJB-A | TAR at FAR=10^-2: 92.3 ± 1.6 -> 96.9 ± 0.8 | Helped |
| Add uniform loss to SphereFace (ablation) | SphereFace · IJB-A | TAR at FAR=10^-3: 88.4 ± 4.2 -> 92.3 ± 1.7 | Helped |
| Add uniform loss to SphereFace (ablation) | SphereFace · IJB-A | Rank-1 identification accuracy: 93.2 ± 1.3 -> 97.9 ± 0.5 | Mixed |
| Add uniform loss to SphereFace (ablation) | SphereFace · YTF | verification rate (%): 96.1 -> 97.7 | Helped |
| Add uniform loss to SphereFace (ablation) | SphereFace · LFW | verification rate (%): 99.5 -> 99.8 | Helped |
| Add uniform loss to SphereFace (ablation) | SphereFace · MegaFace Challenge I | mean minimum average inter-class distance: 1.13 -> 1.45 | Helped |
| Add uniform loss to SphereFace (ablation) | SphereFace · MegaFace Challenge I | variance of minimum average inter-class distances: 0.10 -> 0.06 | Helped |
| Add uniform loss to ArcFace (ablation) | ArcFace · MegaFace Challenge I | Rank-1 accuracy (%): 79.14 -> 81.46 | Helped |
| Add uniform loss to CosFace (ablation) | CosFace · MegaFace Challenge I | Rank-1 accuracy (%): 81.59 -> 83.53 | Helped |
| Vary the A-Softmax margin parameter m (ablation) | A-Softmax · YTF and LFW training experiments | standard deviation of minimum average inter-class distances: 0.02 vs. 0.04 on YTF; 0.05 vs. 0.13 on LFW | Hurt |
Closest papers by meaning
- SphereFace: Deep Hypersphere Embedding for Face Recognition (CVPR 2017 · similarity 0.88)
- ArcFace: Additive Angular Margin Loss for Deep Face Recognition (CVPR 2019 · similarity 0.88)
- RegularFace: Deep Face Recognition via Exclusive Regularization (CVPR 2019 · similarity 0.87)
- CosFace: Large Margin Cosine Loss for Deep Face Recognition (CVPR 2018 · similarity 0.86)
- UniTSFace: Unified Threshold Integrated Sample-to-Sample Loss for Face Recognition (NeurIPS 2023 · similarity 0.86)
- HyperFace: Generating Synthetic Face Recognition Datasets by Exploring Face Embedding Hypersphere (ICLR 2025 · 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.