BriefGPT.xyz
Feb, 2024
探索篡改检测中的显著性偏见
Exploring Saliency Bias in Manipulation Detection
HTML
PDF
Joshua Krinsky, Alan Bettis, Qiuyu Tang, Daniel Moreira, Aparna Bharati
TL;DR
通过分析流行的图像操纵数据集中的视觉和语义显著趋势及其对检测的影响，我们提出了一个框架，以促进语义感知的法庭取证方法的进一步发展，从而了解视觉误导现象。
Abstract
The social media-fuelled explosion of
fake news
and
misinformation
supported by tampered images has led to growth in the development of models and datasets for
→