中文
相关论文

相关论文: All but One: Surgical Concept Erasing with Model P…

200 篇论文

A significant research effort is focused on exploiting the amazing capacities of pretrained diffusion models for the editing of images.They either finetune the model, or invert the image in the latent space of the pretrained model. However,…

计算机视觉与模式识别 · 计算机科学 2024-12-09 Senmao Li , Joost van de Weijer , Taihang Hu , Fahad Shahbaz Khan , Qibin Hou , Yaxing Wang , Jian Yang , Ming-Ming Cheng

Recent advancements in large-scale generative models have enabled the creation of high-quality images and videos, but have also raised significant safety concerns regarding the generation of unsafe content. To mitigate this, concept erasure…

计算机视觉与模式识别 · 计算机科学 2025-11-26 Viet Nguyen , Vishal M. Patel

Erasing harmful or proprietary concepts from powerful text to image generators is an emerging safety requirement, yet current "concept erasure" techniques either collapse image quality, rely on brittle adversarial losses, or demand…

机器学习 · 计算机科学 2025-11-11 Abhiram Kusumba , Maitreya Patel , Kyle Min , Changhoon Kim , Chitta Baral , Yezhou Yang

Generating a coherent sequence of images that tells a visual story, using text-to-image diffusion models, often faces the critical challenge of maintaining subject consistency across all story scenes. Existing approaches, which typically…

计算机视觉与模式识别 · 计算机科学 2025-08-07 Gopalji Gaur , Mohammadreza Zolfaghari , Thomas Brox

Diffusion models, such as Stable Diffusion, have shown incredible performance on text-to-image generation. Since text-to-image generation often requires models to generate visual concepts with fine-grained details and attributes specified…

计算机视觉与模式识别 · 计算机科学 2024-04-26 Xuehai He , Weixi Feng , Tsu-Jui Fu , Varun Jampani , Arjun Akula , Pradyumna Narayana , Sugato Basu , William Yang Wang , Xin Eric Wang

Concept erasure aims to remove harmful, inappropriate, or copyrighted content from text-to-image diffusion models while preserving non-target semantics. However, existing methods either rely on costly fine-tuning or apply coarse semantic…

计算机视觉与模式识别 · 计算机科学 2026-04-14 Ning Han , Zhenyu Ge , Feng Han , Yuhua Sun , Chengqing Li , Jingjing Chen

Text-to-image diffusion models have emerged as powerful tools for high-quality image generation and editing. Many existing approaches rely on text prompts as editing guidance. However, these methods are constrained by the need for manual…

计算机视觉与模式识别 · 计算机科学 2025-05-21 Yuanyuan Chang , Yinghua Yao , Tao Qin , Mengmeng Wang , Ivor Tsang , Guang Dai

Concept erasure helps stop diffusion models (DMs) from generating harmful content; but current methods face robustness retention trade off. Robustness means the model fine-tuned by concept erasure methods resists reactivation of erased…

机器学习 · 计算机科学 2026-02-16 Fengpeng Li , Kemou Li , Qizhou Wang , Bo Han , Jiantao Zhou

Personalized text-to-image models allow users to generate varied styles of images (specified with a sentence) for an object (specified with a set of reference images). While remarkable results have been achieved using diffusion-based…

计算机视觉与模式识别 · 计算机科学 2024-07-19 Fanyue Wei , Wei Zeng , Zhenyang Li , Dawei Yin , Lixin Duan , Wen Li

Editing images with diffusion models under strict training-free constraints remains a significant challenge. While recent optimisation-based methods achieve strong zero-shot edits from text, they struggle to preserve identity and capture…

计算机视觉与模式识别 · 计算机科学 2026-03-04 Niki Foteinopoulou , Ignas Budvytis , Stephan Liwicki

In this paper, we present DesignDiffusion, a simple yet effective framework for the novel task of synthesizing design images from textual descriptions. A primary challenge lies in generating accurate and style-consistent textual and visual…

计算机视觉与模式识别 · 计算机科学 2025-03-04 Zhendong Wang , Jianmin Bao , Shuyang Gu , Dong Chen , Wengang Zhou , Houqiang Li

This paper addresses the societal concerns arising from large-scale text-to-image diffusion models for generating potentially harmful or copyrighted content. Existing models rely heavily on internet-crawled data, wherein problematic…

计算机视觉与模式识别 · 计算机科学 2024-08-01 Sanghyun Kim , Seohyeon Jung , Balhae Kim , Moonseok Choi , Jinwoo Shin , Juho Lee

Diffusion models have dramatically advanced text-to-image generation in recent years, translating abstract concepts into high-fidelity images with remarkable ease. In this work, we examine whether they can also blend distinct concepts,…

计算机视觉与模式识别 · 计算机科学 2025-07-01 Lorenzo Olearo , Giorgio Longari , Alessandro Raganato , Rafael Peñaloza , Simone Melzi

Concept unlearning in diffusion models is hampered by feature splitting, where concepts are distributed across many latent features, making their removal challenging and computationally expensive. We introduce SAEmnesia, a supervised sparse…

计算机视觉与模式识别 · 计算机科学 2025-12-01 Enrico Cassano , Riccardo Renzulli , Marco Nurisso , Mirko Zaffaroni , Alan Perotti , Marco Grangetto

Recent advancements in diffusion models have significantly impacted the trajectory of generative machine learning research, with many adopting the strategy of fine-tuning pre-trained models using domain-specific text-to-image datasets.…

计算机视觉与模式识别 · 计算机科学 2023-12-21 Mischa Dombrowski , Hadrien Reynaud , Johanna P. Müller , Matthew Baugh , Bernhard Kainz

Text-to-image models, such as Stable Diffusion (SD), undergo iterative updates to improve image quality and address concerns such as safety. Improvements in image quality are straightforward to assess. However, how model updates resolve…

密码学与安全 · 计算机科学 2024-09-02 Yixin Wu , Yun Shen , Michael Backes , Yang Zhang

Recent works demonstrate a remarkable ability to customize text-to-image diffusion models while only providing a few example images. What happens if you try to customize such models using multiple, fine-grained concepts in a sequential…

计算机视觉与模式识别 · 计算机科学 2024-05-03 James Seale Smith , Yen-Chang Hsu , Lingyu Zhang , Ting Hua , Zsolt Kira , Yilin Shen , Hongxia Jin

The proliferation of text-to-image diffusion models has raised significant privacy and security concerns, particularly regarding the generation of copyrighted or harmful images. In response, concept erasure (defense) methods have been…

机器学习 · 计算机科学 2025-10-06 Alex D. Richardson , Kaicheng Zhang , Lucas Beerens , Dongdong Chen

Image synthesis approaches, e.g., generative adversarial networks, have been popular as a form of data augmentation in medical image analysis tasks. It is primarily beneficial to overcome the shortage of publicly accessible data and…

计算机视觉与模式识别 · 计算机科学 2023-10-05 Shiyi Du , Xiaosong Wang , Yongyi Lu , Yuyin Zhou , Shaoting Zhang , Alan Yuille , Kang Li , Zongwei Zhou

Large-scale text-to-image (T2I) diffusion models have achieved remarkable generative performance about various concepts. With the limitation of privacy and safety in practice, the generative capability concerning NSFW (Not Safe For Work)…

计算机视觉与模式识别 · 计算机科学 2025-10-01 Jiahang Tu , Qian Feng , Jiahua Dong , Hanbin Zhao , Chao Zhang , Nicu Sebe , Hui Qian