Research map / Label-efficient and robust learning / Recognition and long-tail
RDCFace: Radial Distortion Correction 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 |
|---|---|---|---|
| jointly learn distortion correction, alignment, and recognition from identity labels | RDCFace · D-LFW | verification accuracy (%): 98.27 (-1.51) -> 99.78 (-0.02) | Helped |
| jointly learn distortion correction, alignment, and recognition from identity labels | RDCFace · D-CFP-FP | verification accuracy (%): 63.30 (-34.56) -> 95.30 (-1.32) | Helped |
| jointly learn distortion correction, alignment, and recognition from identity labels | RDCFace · D-YTF | verification accuracy (%): 84.68 (-12.5) -> 96.98 (-0.12) | Helped |
| use distorted-data augmentation without a correction module | Baseline and Align · IMDB-Face test set, Random k | TAR@FAR=1e-2 (%): 45.68 (-37.76) -> 77.27 (-3.43); 83.44 -> 80.70 | Mixed |
| add the inverted foveal layer (ablation) | RDCFace correction network · IMDB-Face test set | distortion coefficient MAE: 0.283 -> 0.241 | Helped |
| add the edge enhancement layer (ablation) | RDCFace correction network · IMDB-Face test set | distortion coefficient MAE: 0.283 -> 0.265 | Helped |
| add the re-correction loss (ablation) | RDCFace correction network · IMDB-Face test set | distortion coefficient MAE: 0.283 -> 0.227 | Helped |
| add the inverted foveal layer, edge enhancement, and re-correction loss (ablation) | RDCFace correction network · IMDB-Face test set | distortion coefficient MAE: 0.283 -> 0.215 | Helped |
| apply correction, alignment, and recognition jointly | RDCFace · RadialFace | mean verification accuracy (%): 93.67% -> 96.67% | Helped |
Closest papers by meaning
- ArcFace: Additive Angular Margin Loss for Deep Face Recognition (CVPR 2019 · similarity 0.85)
- Direct Shape Regression Networks for End-to-End Face Alignment (CVPR 2018 · similarity 0.84)
- Vec2Face: Scaling Face Dataset Generation with Loosely Constrained Vectors (ICLR 2025 · similarity 0.84)
- WebFace260M: A Benchmark Unveiling the Power of Million-Scale Deep Face Recognition (CVPR 2021 · similarity 0.83)
- Discriminative Invariant Kernel Features: A Bells-and-Whistles-Free Approach to Unsupervised Face Recognition and Pose Estimation (CVPR 2016 · similarity 0.83)
- ControlFace: Harnessing Facial Parametric Control for Face Rigging (CVPR 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.