BriefGPT.xyz
Nov, 2016
具有上下文条件生成对抗网络的半监督学习
Semi-Supervised Learning with Context-Conditional Generative Adversarial Networks
HTML
PDF
Remi Denton, Sam Gross, Rob Fergus
TL;DR
本论文介绍了一种基于对抗损失的图像修复半监督学习方法,通过生成器填充图像中的孔洞,再用鉴别器判断图像的真实性,以实现对鉴别器的有监督训练,并在STL-10和PASCAL数据集上表现出与现有方法相当甚至更好的性能。
Abstract
We introduce a simple
semi-supervised learning
approach for images based on
in-painting
using an
adversarial loss
. Images with random patc
→