Research map / Label-efficient and robust learning / Recognition and long-tail
Adversarial Discriminative Heterogeneous Face Recognition
Read the paperCheck 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 |
|---|---|---|---|
| Add feature adversarial loss (ablation) | ADFL · CASIA NIR-VIS 2.0 | Rank-1 acc.(%): 96.56±0.63 -> 97.81±0.29 | Helped |
| Add class-wise variance discrepancy loss (ablation) | ADFL · CASIA NIR-VIS 2.0 | Rank-1 acc.(%): 97.34±0.53 -> 97.81±0.29 | Helped |
| Combine cross-spectral hallucination with ADFL | ADFL · CASIA NIR-VIS 2.0 | Rank-1 acc.(%): 97.81±0.29 -> 98.15±0.34 | Helped |
| Add feature adversarial loss (ablation) | ADFL · BUAA-VisNir | Rank-1: 94.9 -> 95.2 | Helped |
| Add feature adversarial loss (ablation) | ADFL · Oulu-CASIA NIR-VIS | Rank-1: 92.7 -> 95.5 | Helped |
| Add feature adversarial loss (ablation) | ADFL · Oulu-CASIA NIR-VIS | FAR=1%: 83.5 -> 83.0 | Hurt |
| Add class-wise variance discrepancy loss (ablation) | ADFL · Oulu-CASIA NIR-VIS | FAR=0.1%: 55.0 -> 60.7 | Helped |
| Add class-wise variance discrepancy loss (ablation) | ADFL · BUAA-VisNir | FAR=0.1%: 83.9 -> 88.0 | Helped |
| Add feature adversarial loss (ablation) | ADFL · BUAA-VisNir | FAR=0.1%: 83.9 -> 88.0 | Helped |
| Add class-wise variance discrepancy loss (ablation) | ADFL · Oulu-CASIA NIR-VIS | Rank-1: 93.1 -> 95.5 | Helped |
Closest papers by meaning
- Escaping Data Scarcity for High-Resolution Heterogeneous Face Hallucination (CVPR 2022 · similarity 0.89)
- Cross-Spectral Face Hallucination via Disentangling Independent Factors (CVPR 2020 · similarity 0.88)
- Not Afraid of the Dark: NIR-VIS Face Recognition via Cross-Spectral Hallucination and Low-Rank Embedding (CVPR 2017 · similarity 0.87)
- Residual Compensation Networks for Heterogeneous Face Recognition (AAAI 2019 · similarity 0.86)
- UMFN: Unified Multi-Domain Face Normalization for Joint Cross-domain Prototype Learning and Heterogeneous Face Recognition (CVPR 2025 · similarity 0.85)
- Masked Face Recognition with Generative-to-Discriminative Representations (ICML 2024 · similarity 0.83)
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.