Forget-Me-Not: Learning to Forget in Text-to-Image Diffusion Models

By Eric J. Zhang, Kai Wang, Xingqian Xu, Zhangyang Wang, Humphrey Shi. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2023

339 citations recorded by the source corpus.

The significant advances in applications of text-to-image generation models have prompted the demand of a post-hoc adaptation algorithms that can efficiently remove unwanted concepts (e.g. privacy, copyright, and safety) from a pretrained models with minimal influence on the existing knowledge system learned from pretraining. Existing methods mainly resort to explicitly finetuning unwanted concepts to be some alternatives such as their hypernyms or antonyms. Essentially, they are modifying the knowledge system of pretrained models by replacing unwanted to be something arbitrarily defined by user. Furthermore, these methods require hundreds of optimization steps, as they solely rely on denoising loss used for pretraining. To address above challenges, we propose Forget-Me-Not, a model-centric and efficient solution designed to remove identities, objects, or styles from a well-configured text-to-image model in as little as 30 seconds, without significantly impairing its ability to generate other content. In contrast to existing methods, we introduce attention re-steering loss to redirect model’s generation from unwanted concepts to those are learned during pretraining, rather than being user-defined. Furthermore, our method offers two practical extensions: a) removal of potentially harmful or NSFW content, and b) enhancement of model accuracy, inclusion and diversity through concept correction and disentanglement.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/d9b95937934d7291b7c253b28b6c9aaee033c91d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. David Bau
2. Yonatan Belinkov
3. Yanzhu Liu
4. Adams Wai-Kin Kong
5. Leyang Li
6. Zilan Wang
7. Shilin Lu
8. Yihua Zhang

Citations in the corpus

• MACE: Mass Concept Erasure in Diffusion Models (2024, methodology): “Additionally, it is noteworthy that while meth-ods like FMN [71] and AC [30] are proficient in removing specific features of a subject, they fall short in completely eradicating the subject’s generation.”
• Unified Concept Editing in Diffusion Models (2023, methodology): “Another approach adjusts a model after training is complete, deleting an undesired concept by modifying model weights [13, 15, 20, 21, 46].”
• Rethinking machine unlearning for large language models (2024, neutral): “…of MU has expanded to encompass diverse domains, such as image classification Ginart et al. (2019); Golatkar et al. (2020) In the literature, ‘exact’ unlearning, which involves retraining the model from scratch after removing specific training data points, is often considered the gold standard.”
• SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation (2023, background): “Instead, it was initially termed as ‘learning to forget’ or ‘concept erasing’ (Schramowski et al., 2023; Gandikota et al., 2023; Zhang et al., 2023a).”