Research map / Image generation / GANs
CIAGAN: Conditional Identity Anonymization Generative Adversarial Networks
Check 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 |
|---|---|---|---|
| use CIAGAN instead of Gafni et al. | Face anonymization · LFW | True acceptance rate (VGGFace2): 0.038 ± 0.015 -> 0.034 ± 0.016 | Helped |
| use CIAGAN instead of Gafni et al. | Face anonymization · LFW | True acceptance rate (CASIA): 0.035 ± 0.011 -> 0.019 ± 0.008 | Helped |
| use CIAGAN rather than pixelization or blurring | Face anonymization · CelebA | Detection (Dlib): 97.8 | Helped |
| replace the siamese identity discriminator with a classification network (ablation) | CIAGAN · CelebA | Detection: 99.9 -> 64.6 | Hurt |
| replace the siamese identity discriminator with a classification network (ablation) | CIAGAN · CelebA | FID: 2.1 -> 63.2 | Hurt |
| replace the siamese identity discriminator with a classification network (ablation) | CIAGAN · CelebA | Recall@1: 1.3 -> 0.4 | Helped |
| use full face images instead of landmarks as generator input (ablation) | CIAGAN · CelebA | Detection: 99.9 -> 98.3 | Hurt |
| use full face images instead of landmarks as generator input (ablation) | CIAGAN · CelebA | FID: 2.1 -> 6.5 | Hurt |
Closest papers by meaning
- Attribute-Preserving Face Dataset Anonymization via Latent Code Optimization (CVPR 2023 · similarity 0.86)
- Robust Conditional Generative Adversarial Networks (ICLR 2019 · similarity 0.85)
- Balancing Privacy and Performance: A Many-in-One Approach for Image Anonymization (AAAI 2025 · similarity 0.84)
- ToonerGAN: Reinforcing GANs for Obfuscating Automated Facial Indexing (CVPR 2024 · similarity 0.84)
- IDProtector: An Adversarial Noise Encoder to Protect Against ID-Preserving Image Generation (CVPR 2025 · similarity 0.83)
- Attack Deterministic Conditional Image Generative Models for Diverse and Controllable Generation (AAAI 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.