Vid-Freeze: Protecting Images from Malicious Image-to-Video Generation via Temporal Freezing

The rapid progress of image-to-video (I2V) generation models has introduced significant risks, enabling video synthesis from static images and facilitating deceptive or malicious content creation. While prior defenses such as I2VGuard attempt to immunize images, effective and principled protection to block motion remains underexplored. In this work, we introduce Vid-Freeze - a novel attention-suppressing adversarial attack that adds carefully crafted adversarial perturbations to images. Our method explicitly targets the attention mechanism of I2V models, completely disrupting motion synthesis while preserving semantic fidelity of the input image. The resulting immunized images generate stand-still or near-static videos, effectively blocking malicious content creation. Our experiments demonstrate the impressive protection provided by the proposed approach, highlighting the importance of attention attacks as a promising direction for robust and proactive defenses against misuse of I2V generation models.

Key Contributions

• Principled layer-selection strategy identifying the most vulnerable attention layers in I2V diffusion transformers (CogVideoX, SVD)
• Attention suppression loss that collapses generated video onto the static input frame, achieving complete temporal freezing
• Demonstrates temporal freezing with minimal perturbation budgets — sometimes modifying as few as 2 pixels — outperforming prior I2VGuard defense

🛡️ Threat Analysis

Details

Similar Papers