Implicit Concept Removal of Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Zhili, Chen, Kai, Zhang, Yifan, Han, Jianhua, Hong, Lanqing, Xu, Hang, Li, Zhenguo, Yeung, Dit-Yan, Kwok, James
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912062307827712
author Liu, Zhili
Chen, Kai
Zhang, Yifan
Han, Jianhua
Hong, Lanqing
Xu, Hang
Li, Zhenguo
Yeung, Dit-Yan
Kwok, James
author_facet Liu, Zhili
Chen, Kai
Zhang, Yifan
Han, Jianhua
Hong, Lanqing
Xu, Hang
Li, Zhenguo
Yeung, Dit-Yan
Kwok, James
contents Text-to-image (T2I) diffusion models often inadvertently generate unwanted concepts such as watermarks and unsafe images. These concepts, termed as the "implicit concepts", could be unintentionally learned during training and then be generated uncontrollably during inference. Existing removal methods still struggle to eliminate implicit concepts primarily due to their dependency on the model's ability to recognize concepts it actually can not discern. To address this, we utilize the intrinsic geometric characteristics of implicit concepts and present the Geom-Erasing, a novel concept removal method based on the geometric-driven control. Specifically, once an unwanted implicit concept is identified, we integrate the existence and geometric information of the concept into the text prompts with the help of an accessible classifier or detector model. Subsequently, the model is optimized to identify and disentangle this information, which is then adopted as negative prompts during generation. Moreover, we introduce the Implicit Concept Dataset (ICD), a novel image-text dataset imbued with three typical implicit concepts (i.e., QR codes, watermarks, and text), reflecting real-life situations where implicit concepts are easily injected. Geom-Erasing effectively mitigates the generation of implicit concepts, achieving the state-of-the-art results on the Inappropriate Image Prompts (I2P) and our challenging Implicit Concept Dataset (ICD) benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2310_05873
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Implicit Concept Removal of Diffusion Models
Liu, Zhili
Chen, Kai
Zhang, Yifan
Han, Jianhua
Hong, Lanqing
Xu, Hang
Li, Zhenguo
Yeung, Dit-Yan
Kwok, James
Computer Vision and Pattern Recognition
Text-to-image (T2I) diffusion models often inadvertently generate unwanted concepts such as watermarks and unsafe images. These concepts, termed as the "implicit concepts", could be unintentionally learned during training and then be generated uncontrollably during inference. Existing removal methods still struggle to eliminate implicit concepts primarily due to their dependency on the model's ability to recognize concepts it actually can not discern. To address this, we utilize the intrinsic geometric characteristics of implicit concepts and present the Geom-Erasing, a novel concept removal method based on the geometric-driven control. Specifically, once an unwanted implicit concept is identified, we integrate the existence and geometric information of the concept into the text prompts with the help of an accessible classifier or detector model. Subsequently, the model is optimized to identify and disentangle this information, which is then adopted as negative prompts during generation. Moreover, we introduce the Implicit Concept Dataset (ICD), a novel image-text dataset imbued with three typical implicit concepts (i.e., QR codes, watermarks, and text), reflecting real-life situations where implicit concepts are easily injected. Geom-Erasing effectively mitigates the generation of implicit concepts, achieving the state-of-the-art results on the Inappropriate Image Prompts (I2P) and our challenging Implicit Concept Dataset (ICD) benchmarks.
title Implicit Concept Removal of Diffusion Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2310.05873