Research map / Label-efficient and robust learning / Recognition and long-tail
Learning To Learn Across Diverse Data Biases in Deep Face Recognition
Chang Liu, Xiang Yu, Yi-Hsuan Tsai, Masoud Faraki, Ramin Moslemi, Manmohan Chandraker, Yun Fu
Check 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 occlusion-aware margin weighting (ablation) | CosFace · OC-LFW | accuracy: 94.41 -> 94.52 | Improved |
| Add all modeled variation factors (ablation) | CosFace · OC-LFW | accuracy: 94.41 -> 94.83 | Improved |
| Add head-pose-aware margin weighting (ablation) | CosFace · CFP-FP | accuracy: 98.16 -> 98.35 | Improved |
| Add all modeled variation factors (ablation) | CosFace · CFP-FP | accuracy: 98.16 -> 98.41 | Improved |
| Add all modeled variation factors (ablation) | CosFace · RFW average | accuracy: 97.94 -> 98.32 | Improved |
| Add all modeled variation factors (ablation) | CosFace · RFW | bias: 0.67 -> 0.51 | Improved |
| Add blur-aware margin weighting (ablation) | CosFace · IJB-A (Vrf), FAR@0.001% | verification rate: 93.22 -> 95.58 | Improved |
| Combine ethnicity and pose variation margins (ablation) | CosFace · IJB-A (Vrf), FAR@0.01% | verification rate: 97.13 -> 97.20 | No clear effect |
| Add blur variation margin to ethnicity and pose (ablation) | CosFace · IJB-A (Vrf), FAR@0.01% | verification rate: 97.20 -> 97.45 | Improved |
| Add all modeled variation factors (ablation) | CosFace · IJB-A (Vrf), FAR@0.01% | verification rate: 97.13 -> 97.46 | Improved |
| Replace CosFace with MvCoM-CosFace | CosFace · LFW | accuracy: 99.73 -> 99.80 | No clear effect |
| Replace CosFace with MvCoM-CosFace | CosFace · MegaFace1 uncleaned, Rank1 | Rank1: 80.03 -> 81.30 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 4 on most shared results.
- CosFace — CosFace: Large Margin Cosine Loss for Deep Face Recognition (CVPR 2018) · better on 3 of 3 shared results
- ArcFace — ArcFace: Additive Angular Margin Loss for Deep Face Recognition (CVPR 2019) · better on 1 of 2 shared results
- MagFace — MagFace: A Universal Representation for Face Recognition and Quality Assessment (CVPR 2021) · better on 0 of 1 shared result
- CurricularFace — CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognition (CVPR 2020) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- CosFace: Large Margin Cosine Loss for Deep Face Recognition (CVPR 2018 · similarity 0.87)
- Noise-Tolerant Paradigm for Training Face Recognition CNNs (CVPR 2019 · similarity 0.87)
- AdaptiveFace: Adaptive Margin and Sampling for Face Recognition (CVPR 2019 · similarity 0.87)
- Deep Imbalanced Learning for Face Recognition and Attribute Prediction (TPAMI 2019 · similarity 0.86)
- Mitigating Bias in Face Recognition Using Skewness-Aware Reinforcement Learning (CVPR 2020 · similarity 0.86)
- Meta Balanced Network for Fair Face Recognition (TPAMI 2021 · similarity 0.85)
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.
