Conditional Image Synthesis with Auxiliary Classifier GANs
Conditional Image Synthesis with Auxiliary Classifier GANs
复制标题
DOI：
--
复制
发表时间：
2016-10
影响因子：
5.6
通讯作者：
Augustus Odena;C. Olah;Jonathon Shlens
中科院分区：
文献类型：
--
作者：
Augustus Odena;C. Olah;Jonathon Shlens
在本文中，我们介绍了改进图像合成生成对抗网络（GAN）训练的新方法。我们构建了一种采用标签调节的 GAN 变体，可生成具有全局一致性的 128 x 128 分辨率图像样本。我们扩展了之前的图像质量评估工作，提供了两种新的分析来评估类条件图像合成模型中样本的可辨别性和多样性。这些分析表明，高分辨率样本提供了低分辨率样本中不存在的类别信息。在 1000 个 ImageNet 类中，128 x 128 样本的可辨别性是人工调整大小的 32 x 32 样本的两倍多。此外，84.7% 的类别的样本表现出与真实 ImageNet 数据相当的多样性。
In this paper we introduce new methods for the improved training of generative adversarial networks (GANs) for image synthesis. We construct a variant of GANs employing label conditioning that results in 128 x 128 resolution image samples exhibiting global coherence. We expand on previous work for image quality assessment to provide two new analyses for assessing the discriminability and diversity of samples from class-conditional image synthesis models. These analyses demonstrate that high resolution samples provide class information not present in low resolution samples. Across 1000 ImageNet classes, 128 x 128 samples are more than twice as discriminable as artificially resized 32 x 32 samples. In addition, 84.7% of the classes have samples exhibiting diversity comparable to real ImageNet data.