Research map / Label-efficient and robust learning / Recognition and long-tail
FM2u-Net: Face Morphological Multi-Branch Network for Makeup-Invariant Face Verification
Wenxuan Wang, Yanwei Fu, Xuelin Qian, Yu-Gang Jiang, Qi Tian, Xiangyang Xue
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 |
|---|---|---|---|
| Use FM²u-Net instead of the strongest listed baseline | Makeup face verification · M-501 | accuracy: 94.80 -> 98.12 | Improved |
| Use FM²u-Net instead of the strongest listed baseline | Makeup face verification · M-203 | accuracy: 92.30 -> 94.34 | Improved |
| Use FM²u-Net instead of the strongest listed baseline | Makeup face verification · FAM | accuracy: 88.73 -> 90.43 | Improved |
| Use FM²u-Net instead of the strongest listed baseline | Makeup face verification · Extended | accuracy: 91.92 -> 96.56 | Improved |
| Use FM²u-Net instead of AdaCos | General face verification · LFW+ | accuracy: 95.12 -> 97.89 | Improved |
| Add FM-Net-generated synthetic makeup images (ablation) | FM²u-Net · M-501 | accuracy: 95.22 -> 98.12 | Improved |
| Add FM-Net-generated synthetic makeup images (ablation) | FM²u-Net · LFW+ | accuracy: 95.34 -> 97.89 | Improved |
| Add AttM-Net in place of LightCNN-29v2 features (ablation) | FM²u-Net · M-501 | accuracy: 95.98 -> 98.12 | Improved |
| Add AttM-Net in place of LightCNN-29v2 features (ablation) | FM²u-Net · LFW+ | accuracy: 96.07 -> 97.89 | Improved |
| Use soft FM-Net face morphology rather than hard copy-and-paste patch replacement (ablation) | FM²u-Net · M-501 | accuracy: 97.14 -> 98.12 | Improved |
| Use FM-Net morphology rather than random-noise patches (ablation) | FM²u-Net · M-501 | accuracy: 96.48 -> 98.12 | Improved |
| Swap the two eyes and mouth rather than randomly chosen patches (ablation) | FM²u-Net · M-501 | accuracy: 96.92 -> 98.12 | Improved |
Closest papers by meaning
- Anti-Makeup: Learning A Bi-Level Adversarial Network for Makeup-Invariant Face Verification (AAAI 2018 · similarity 0.88)
- LADN: Local Adversarial Disentangling Network for Facial Makeup and De-Makeup (ICCV 2019 · similarity 0.86)
- Vec2Face: Scaling Face Dataset Generation with Loosely Constrained Vectors (ICLR 2025 · similarity 0.86)
- UMFN: Unified Multi-Domain Face Normalization for Joint Cross-domain Prototype Learning and Heterogeneous Face Recognition (CVPR 2025 · similarity 0.84)
- Masked Face Recognition with Generative-to-Discriminative Representations (ICML 2024 · similarity 0.84)
- UIFace: Unleashing Inherent Model Capabilities to Enhance Intra-Class Diversity in Synthetic Face Recognition (ICLR 2025 · 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.
