Pith. sign in

REVIEW 11 cited by

Raising the Cost of Malicious AI-Powered Image Editing

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2302.06588 v1 pith:B4STH6QH submitted 2023-02-13 cs.LG

Hadi Salman , Alaa Khaddaj , Guillaume Leclerc , Andrew Ilyas , Aleksander Madry This is my paper · ORCID
classification cs.LG
keywords modelsdiffusionapproacheditingimageimagesimmunizationmake
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We present an approach to mitigating the risks of malicious image editing posed by large diffusion models. The key idea is to immunize images so as to make them resistant to manipulation by these models. This immunization relies on injection of imperceptible adversarial perturbations designed to disrupt the operation of the targeted diffusion models, forcing them to generate unrealistic images. We provide two methods for crafting such perturbations, and then demonstrate their efficacy. Finally, we discuss a policy component necessary to make our approach fully effective and practical -- one that involves the organizations developing diffusion models, rather than individual users, to implement (and support) the immunization process.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 11 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. I2VShield: An Efficient Proactive Defense Framework against DiT-based Image-to-Video Models

    cs.CV 2026-07 conditional novelty 7.0 of 10

    A text-adaptive generator trained to disrupt diffusion-transformer cross-attention shields reference images from video-generation misuse at low online cost.

  2. Cross-Branch Conflict as a Shield: Safeguarding Facial Identities in Unified Multimodal Image Editing

    cs.CV 2026-07 conditional novelty 7.0 of 10

    CCS jointly perturbs the ViT and VAE pathways and reduces their CKA agreement, causing unified multimodal image editors to lose facial identity.

  3. Beyond Invisibility: Learning Robust Visible Watermarks for Stronger Copyright Protection

    cs.LG 2025-06 conditional novelty 7.0 of 10

    HARVIM learns watermark placement to maximize reconstruction error under an inpainting-based removal model, showing modest gains over random watermarks.

  4. IDDM: Identity-Decoupled Personalized Diffusion Models with a Tunable Privacy-Utility Trade-off

    cs.CV 2026-04 conditional novelty 6.5 of 10

    IDDM immunizes authorized personalized diffusion models so public generations remain high-quality while identity linkability to face recognizers is reduced with a tunable privacy-utility knob.

  5. Defending from GeoLocalization through Adversarial Road Trips

    cs.CV 2026-07 conditional novelty 6.0 of 10

    RoadTrip Attack uses beam search over adaptive geographic intermediate targets to produce stronger, more transferable, lower-visibility adversarial examples against retrieval-based image geolocalizers than PGD, FGSM, ...

  6. SyncBreaker:Stage-Aware Multimodal Adversarial Attacks on Audio-Driven Talking Head Generation

    cs.CV 2026-04 unverdicted novelty 6.0 of 10

    SyncBreaker jointly attacks image and audio streams with Multi-Interval Sampling and Cross-Attention Fooling to degrade speech-driven talking head generation more than single-modality baselines.

  7. Silence is Golden: Leveraging Adversarial Examples to Nullify Audio Control in LDM-based Talking-Head Generation

    cs.GR 2025-06 conditional novelty 6.0 of 10

    Silencer adds a nearly invisible disturbance to portraits that makes LDM-based talking-head models keep the mouth silent, and it survives several image-purification countermeasures.

  8. DECAF: De-Clustering for Adaptive Representational Unlearning

    cs.LG 2026-07 conditional novelty 5.0 of 10

    DECAF is a forget-only unlearning method that adds input noise, suppresses the forget-class probability, and diversifies outputs, achieving 0.10% forget accuracy and 79.4% retain accuracy on CIFAR-10/ResNet-18 while d...

  9. Immunizing Images from Text to Image Editing via Adversarial Cross-Attention

    cs.CV 2025-09 conditional novelty 5.0 of 10

    An imperceptible adversarial noise, computed with a LLaVA caption as a stand-in for the unknown edit prompt, disrupts cross-attention in Stable Diffusion-based editors and makes text-guided edits fail.

  10. Evaluating Adversarial Protections for Diffusion Personalization: A Comprehensive Study

    cs.CV 2025-07 conditional novelty 5.0 of 10

    A unified benchmark of eight perturbation-based protections shows budget-dependent trade-offs between stealth and disruption, with no method winning across all metrics.

  11. Is Perturbation-Based Image Protection Disruptive to Image Editing?

    cs.CV 2025-06 conditional novelty 4.0 of 10

    Perturbation-based protections do not reliably block diffusion editing, and in many cases they increase the edited image's alignment with the guidance prompt.

Pith tools