Research map / Image generation / Editing and style transfer
RPG-Palm: Realistic Pseudo-data Generation for Palmprint Recognition
Lei Shen, Jianlong Jin, Ruixin Zhang, Huaen Li, Kai Zhao, Yingyi Zhang, Jingyun Zhang, Shouhong Ding, Yang Zhao, Wei Jia
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 ID-aware loss to the generator (ablation) | RPG-Palm · Open-set 1:1 | TAR@FAR=1e-6: 0.7210 -> 0.8382 | Improved |
| add conditional modulation generator (CAdaIN) instead of plain UNet (ablation) | RPG-Palm · Open-set 1:1 | TAR@FAR=1e-6: 0.8382 -> 0.8831 | Improved |
| add improved Bezier palm creases synthesis (ablation) | RPG-Palm · Open-set 1:1 | TAR@FAR=1e-6: 0.8831 -> 0.8946 | Improved |
| use RPG-Palm generator instead of pix2pixHD (ablation) | synthetic pretraining generator · Open-set 1:1 | FID: 97.5801 -> 16.4762 | Improved |
| use BicycleGAN instead of RPG-Palm generator (ablation) | synthetic pretraining generator · Open-set 1:1 | TAR@FAR=1e-6: 0.8946 -> 0.7254 | Hurt |
| RPG-Palm pretraining vs BezierPalm pretraining with few real IDs (80) | MobileFaceNet · Limited real IDs, open-set 1:1 | TAR@FAR=1e-6: 0.3743 -> 0.5824 | Improved |
| RPG-Palm pretraining with 160 real IDs vs ArcFace with 1600 real IDs | MobileFaceNet · Limited real IDs, open-set 1:1 | TAR@FAR=1e-6: 0.7049 -> 0.7246 | Improved |
| RPG-Palm pretraining vs BezierPalm on million-scale data (ResNet50) | ResNet50 · Million-scale internal dataset | TAR@FAR=1e-6: 0.9996 -> 0.9998 | No clear effect |
| increase synthetic identities and samples per identity (ablation) | RPG-Palm · Open-set 1:1 | TAR@FAR=1e-6 | Improved |
| RPG-Palm vs BezierPalm under 1:3 split | MobileFaceNet · Open-set 1:3 | TAR@FAR=1e-6: 0.7012 -> 0.8485 | Improved |
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 0.91)
- PVTree: Realistic and Controllable Palm Vein Generation for Recognition Tasks (AAAI 2025 · similarity 0.85)
- Analyzing the Synthetic-to-Real Domain Gap in 3D Hand Pose Estimation (CVPR 2025 · similarity 0.79)
- HanDiffuser: Text-to-Image Generation With Realistic Hand Appearances (CVPR 2024 · similarity 0.79)
- ManiFPT: Defining and Analyzing Fingerprints of Generative Models (CVPR 2024 · similarity 0.78)
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.
