Research map / Label-efficient and robust learning / Recognition and long-tail
BézierPalm: A Free Lunch for Palmprint Recognition
Kai Zhao, Lei Shen, Yingyi Zhang, Chuhan Zhou, Tao Wang, Ruixin Zhang, Shouhong Ding, Wei Jia, Wei Shen
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 synthetic Bezier palmprint pretraining | ArcFace MobileFaceNet · Open-set 1:1 | TAR@FAR=1e-6: 0.7049 -> 0.8437 | Improved |
| add synthetic Bezier palmprint pretraining | ArcFace ResNet50 · Open-set 1:1 | TAR@FAR=1e-6: 0.7462 -> 0.8956 | Improved |
| add synthetic pretraining | ArcFace MobileFaceNet · Open-set 1:3 | TAR@FAR=1e-6: 0.5825 -> 0.7012 | Improved |
| add synthetic pretraining | ArcFace MobileFaceNet · Closed-set CASIA | EER: 0.009 -> 0.004 | Improved |
| add synthetic pretraining | ArcFace MobileFaceNet · Cross-dataset T->I | TAR@FAR=1e-5: 0.7306 -> 0.9745 | Improved |
| add synthetic pretraining | ArcFace ResNet50 · Million-scale dataset | TAR@FAR=1e-9: 0.9879 -> 0.9911 | Improved |
| add synthetic pretraining | ArcFace MobileFaceNet · 400 training identities | TAR@FAR=1e-6: 0.3320 -> 0.6899 | Improved |
| add principal lines to synthetic samples (ablation) | synthetic pretraining · Open-set 1:1 | TAR@FAR=1e-6: 0.7217 -> 0.7513 | Improved |
| add wrinkles to synthetic samples (ablation) | synthetic pretraining · Open-set 1:1 | TAR@FAR=1e-6: 0.7513 -> 0.8061 | Improved |
| add natural image background to synthetic samples (ablation) | synthetic pretraining · Open-set 1:1 | TAR@FAR=1e-6: 0.8061 -> 0.8437 | Improved |
| use synthetic pretraining instead of ImageNet pretraining (ablation) | MobileFaceNet · Open-set 1:1 | TAR@FAR=1e-6: 0.7256 -> 0.8437 | Improved |
| use synthetic pretraining instead of ImageNet pretraining (ablation) | MobileFaceNet · Open-set 1:1 | TAR@FAR=1e-3: 0.9608 -> 0.9640 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ArcFace — ArcFace: Additive Angular Margin Loss for Deep Face Recognition (CVPR 2019) · better on 15 of 15 shared results
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
- Diff-Palm: Realistic Palmprint Generation with Polynomial Creases and Intra-Class Variation Controllable Diffusion Models (CVPR 2025 · similarity 0.92)
- PCE-Palm: Palm Crease Energy Based Two-Stage Realistic Pseudo-Palmprint Generation (AAAI 2024 · similarity 0.89)
- RPG-Palm: Realistic Pseudo-data Generation for Palmprint Recognition (ICCV 2023 · similarity 0.89)
- FlowPalm: Optical Flow Driven Non-Rigid Deformation for Geometrically Diverse Palmprint Generation (CVPR 2026 · similarity 0.88)
- PVTree: Realistic and Controllable Palm Vein Generation for Recognition Tasks (AAAI 2025 · similarity 0.83)
- Towards Explainable and Unprecedented Accuracy in Matching Challenging Finger Crease Patterns (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.
