Now You See It, Now You Don't - Instant Concept Erasure for Safe Text-to-Image and Video Generation
Shristi Das Biswas, Arani Roy, Kaushik Roy

TL;DR
The paper introduces Instant Concept Erasure (ICE), a training-free, modality-agnostic method for precise and persistent concept removal in text-to-image and video models, enhancing safety with minimal performance impact.
Contribution
ICE presents a novel, analytical approach for one-shot concept erasure that works across T2I and T2V models without retraining or inference overhead.
Findings
Achieves targeted concept removal with minimal collateral damage.
Demonstrates robustness against adversarial red-teaming.
Maintains core generative capabilities post-erasure.
Abstract
Robust concept removal for text-to-image (T2I) and text-to-video (T2V) models is essential for their safe deployment. Existing methods, however, suffer from costly retraining, inference overhead, or vulnerability to adversarial attacks. Crucially, they rarely model the latent semantic overlap between the target erase concept and surrounding content -- causing collateral damage post-erasure -- and even fewer methods work reliably across both T2I and T2V domains. We introduce Instant Concept Erasure (ICE), a training-free, modality-agnostic, one-shot weight modification approach that achieves precise, persistent unlearning with zero overhead. ICE defines erase and preserve subspaces using anisotropic energy-weighted scaling, then explicitly regularises against their intersection using a unique, closed-form overlap projector. We pose a convex and Lipschitz-bounded Spectral Unlearning…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdversarial Robustness in Machine Learning · Generative Adversarial Networks and Image Synthesis · Digital Media Forensic Detection
