While GANs have shown success in realistic image generation, the idea of using GANs for other tasks unrelated to synthesis is underexplored. Do GANs learn meaningful structural parts of objects during their attempt to reproduce those objects? In this work, we test this hypothesis and propose a simple and effective approach based on GANs for semantic part segmentation that requires as few as one label example along with an unlabeled dataset. Our key idea is to leverage a trained GAN to extract pixel-wise representation from the input image and use it as feature vectors for a segmentation network. Our experiments demonstrate that GANs representation is "readily discriminative" and produces surprisingly good results that are comparable to those from supervised baselines trained with significantly more labels. We believe this novel repurposing of GANs underlies a new class of unsupervised representation learning that is applicable to many other tasks. More results are available at https://repurposegans.github.io/.
翻译:虽然GANs在现实的图像生成中表现出了成功,但使用GANs来完成与合成无关的其他任务的想法没有得到充分探讨。 Do GANs在试图复制这些对象的过程中学会了物体中有意义的结构部分。 在这项工作中,我们测试了这一假设,并提出了一个基于GANs的语义分离简单而有效的方法,这需要像一个标签示例一样少的标签和没有标签的数据集。我们的关键想法是利用经过培训的GAN从输入图像中提取像素,并将它用作分解网络的特性矢量。我们的实验表明,GANs的表示方式“明显具有歧视性”,并产生出出出出惊人的好结果,与经过大量标签培训的受监督基线的结果相类似。我们相信,这种对GANs的新重新定位是一个新的非监控的表达学习类别的基础,适用于许多其他任务。更多的结果可在 https://reotectritegans.github.io/上查阅。