DirectEdit: Step-Level Accurate Inversion for Flow-Based Image Editing
DirectEdit: 基于流的图像编辑的逐步骤精确反演
Desong Yang, Mang Ye
机构
*
National Engineering Research Center for Multimedia Software, School of Computer Science, Wuhan University, Wuhan, China(多媒体软件国家工程研究中心,计算机科学学院,武汉大学,中国武汉)
机构
*
Shenzhen Campus of Sun Yat-sen University(中山大学深圳校区)
;
Guangdong Provincial Key Laboratory of Intelligent Information Processing and Shenzhen Key Laboratory of Media Security(广东省智能信息处理重点实验室和深圳媒体安全重点实验室)
;
Shenzhen University of Advanced Technology and Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences(深圳先进技术大学和深圳先进技术研究所,中国科学院)
;
Alibaba Group(阿里巴巴集团)
;
Shenzhen MSU-BIT University(深圳MSU-BIT大学)
Commentsv1:The Forensic Cost of Watermark Removal, accepted at IH&MMSEC 2026, Special Session "Watermarking Across the Lifecycle of Generative Models". v2: extended version, under review
BiasEdit: A Training-Free Bias-Detect-and-Edit Framework for Learning Fair Visual Classifiers
BiasEdit: 一种无需训练的偏差检测与编辑框架,用于学习公平的视觉分类器
Jungwook Seo, Yoonsik Park, Changmin Lee, Sungyong Baik
机构
*
Hanyang University Department of Artificial Intelligence BAIK Lab Seoul South Korea(翰阳大学人工智能系BAIK实验室首尔韩国)
;
Hanyang University Department of Data Science BAIK Lab Seoul South Korea(翰阳大学数据科学系BAIK实验室首尔韩国)
;
Hanyang University Department of Data Science Department of Artificial Intelligence BAIK Lab Seoul South Korea(翰阳大学数据科学系人工智能系BAIK实验室首尔韩国)
;
Hanyang University(翰阳大学)