Research map / Video / Video generation
Anti-I2V: Safeguarding your Photos from Malicious Image-to-video Generation
Duc Vu, Anh Nguyen, Chi Tran, Anh Tran
Read the paperCheck 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 |
|---|---|---|---|
| Perturb in Lab plus frequency space (DSP) instead of RGB (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | Q-Align (video): 0.692 -> 0.610 | Improved |
| Use Lab space alone instead of RGB (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | Q-Align (video): 0.692 -> 0.765 | Hurt |
| Add RGB space to DSP (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | Q-Align (video): 0.610 -> 0.639 | Hurt |
| Add IRC loss to denoising loss (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | ISM: 0.582 -> 0.535 | Improved |
| Add targeted IRA on VAE to denoising loss (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | ISM: 0.582 -> 0.507 | Improved |
| Add targeted IRA on denoiser to denoising loss (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | ISM: 0.582 -> 0.493 | Improved |
| Use targeted instead of untargeted IRA on denoiser (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | ISM: 0.507 -> 0.493 | Improved |
| Add IRC to combined IRA losses (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | ISM: 0.484 -> 0.476 | Improved |
| Add auxiliary LPIPS and CLIP loss (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | ISM: 0.476 -> 0.462 | Improved |
| Raise Lab budget from 16/255 to 32/255 (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | Q-Align (video): 0.562 -> 0.588 | Hurt |
| Lower Lab budget from 16/255 to 8/255 (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | Q-Align (video): 0.562 -> 0.567 | Hurt |
| Apply IRC to last 3 layers instead of all layers after 27th (ablation) | Anti-I2V · CelebV-Text subset, CogVideoX-5B | ISM: 0.458 -> 0.462 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MIST — MIST: Multiple Instance Self-Training Framework for Video Anomaly Detection (CVPR 2021) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- I2VGuard: Safeguarding Images against Misuse in Diffusion-based Image-to-Video Models (CVPR 2025 · similarity 0.94)
- AdvI2I: Adversarial Image Attack on Image-to-Image Diffusion Models (ICML 2025 · similarity 0.89)
- DiffusionGuard: A Robust Defense Against Malicious Diffusion-based Image Editing (ICLR 2025 · similarity 0.89)
- OSV: One Step is Enough for High-Quality Image to Video Generation (CVPR 2025 · similarity 0.88)
- AdvPaint: Protecting Images from Inpainting Manipulation via Adversarial Attention Disruption (ICLR 2025 · similarity 0.88)
- Pixel Is Not a Barrier: An Effective Evasion Attack for Pixel-Domain Diffusion Models (AAAI 2025 · similarity 0.87)
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.
