Research map / Label-efficient and robust learning / Recognition and long-tail
Demodalizing Face Recognition with Synthetic Samples
Zhonghua Zhai, Pengju Yang, Xiaofeng Zhang, Maji Huang, Haijing Cheng, Xuejun Yan, Chunmao Wang, Shiliang Pu
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 |
|---|---|---|---|
| Use synthetic samples with explicit identity/non-identity disentangling | CosFace · IJB-A | TAR@FAR=1e-4: 0.61718 -> 0.86662 | Improved |
| Use synthetic samples with explicit identity/non-identity disentangling | CosFace · IJB-C | TAR@FAR=1e-5: 0.55689 -> 0.84819 | Improved |
| Use synthetic samples with explicit identity/non-identity disentangling | CosFace · CQ-IJB-C | TAR@FAR=1e-5: 0.26264 -> 0.59166 | Improved |
| Use synthetic samples with explicit identity/non-identity disentangling | CosFace · CP-IJB-C | TAR@FAR=1e-5: 0.10385 -> 0.5393 | Improved |
| Use synthetic samples with implicit meta-learning demodalization | CosFace · IJB-A | TAR@FAR=1e-3: 0.85133 -> 0.91923 | Improved |
| Use synthetic samples with implicit meta-learning demodalization | CosFace · CP-IJB-C | TAR@FAR=1e-5: 0.10385 -> 0.157767 | Improved |
| Use synthetic samples with feature-filter demodalization | CosFace · CQ-IJB-C | TAR@FAR=1e-5: 0.26264 -> 0.37303 | Improved |
| Use synthetic samples with feature-filter demodalization | CosFace · CP-IJB-C | TAR@FAR=1e-5: 0.10385 -> 0.17101 | Improved |
| Train with synthetic samples using DMFR rather than without synthetic samples | CosFace · LFW | accuracy: 0.9975 -> 0.9972 | No clear effect |
| Train with synthetic samples using DMFR rather than without synthetic samples | CosFace · MegaFace | accuracy: 0.7893 -> 0.7660 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 on most shared results.
- ArcFace — ArcFace: Additive Angular Margin Loss for Deep Face Recognition (CVPR 2019) · better on 0 of 1 shared result
- CosFace — CosFace: Large Margin Cosine Loss for Deep Face Recognition (CVPR 2018) · 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
- UIFace: Unleashing Inherent Model Capabilities to Enhance Intra-Class Diversity in Synthetic Face Recognition (ICLR 2025 · similarity 0.86)
- Synthetic Face Datasets Generation via Latent Space Exploration from Brownian Identity Diffusion (ICML 2025 · similarity 0.85)
- $\text{ID}^3$: Identity-Preserving-yet-Diversified Diffusion Models for Synthetic Face Recognition (NeurIPS 2024 · similarity 0.85)
- HyperFace: Generating Synthetic Face Recognition Datasets by Exploring Face Embedding Hypersphere (ICLR 2025 · similarity 0.84)
- AugGen: Synthetic Augmentation using Diffusion Models Can Improve Recognition (NeurIPS 2025 · similarity 0.84)
- Real-Fake: Effective Training Data Synthesis Through Distribution Matching (ICLR 2024 · 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.
