Zebin You, Xinyu Zhang, Hanzhong Guo, Jingdong Wang, Chongxuan Li
TL;DR用分布分类任务发现分类器对图像的识别以边缘和高频成分为主,与人类呈现截然不同的观察结果。
Abstract
The ultimate goal of generative models is to characterize the data
distribution perfectly. For image generation, common metrics of visual quality
(e.g., FID), and the truthlikeness of generated images to the human eyes seem
to suggest that we are close to achieving it. However, through