No Way To Steal My Face: Proactive Defense Against Identity-Preserving Personalized Generation
To bridge this gap, we propose IDGuardian-the first generalizable and model-agnostic identity protection framework capable of defending against both training-based and training-free personalization methods.
ProofPaper ↗
Key points
- Recent advances in diffusion models have enabled high-fidelity, identity-preserving image generation for personalized applications such as digital avatars and virtual try-on systems.
- IDGuardian abstracts the personalization process into two critical stages: identity extraction and identity injection.
- It then introduces crafted adversarial perturbations to simultaneously disrupt both stages.
- Extensive experiments show that IDGuardian effectively protects identity across various personalization pipelines and model architectures, while remaining robust to post-processing, adaptive attacks, and cross-dataset generalization.
Sources (1)
- [1]No Way To Steal My Face: Proactive Defense Against Identity-Preserving Personalized GenerationarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 14, 07:43 AM
To bridge this gap, we propose IDGuardian-the first generalizable and model-agnostic identity protection framework capable of defending against both training-based and training-free personalization methods.
Recent advances in diffusion models have enabled high-fidelity, identity-preserving image generation for personalized applications such as digital avatars and virtual try-on systems.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 14, 2026SonoWorld: Towards Interactive Pose-Controllable Ultrasound World Models
- Oct 14, 2026ReaDy: Recurrent States as Dynamic Context for Speculative Drafting in Hybrid Linear Attention Large Language Models
- Oct 13, 2026VigenFlow: FLUX.1-schnell, FLUX.2-klein-4B and Z-Image-Turbo running locally on the AMD Ryzen AI NPU, no discrete GPU needed (Windows + Ubuntu)
- Oct 13, 2026Discovering Undisclosed Concepts in LoRA Adapters for Diffusion-Based AIGC Models
- Oct 13, 2026LinMorph: Learning to Morph Weights for Efficient Video DiT Linearization
- Oct 13, 2026Risk-Averse Inference-Time Diffusion Model Alignment with Finite-Sample Guarantees