FineGAN: Unsupervised Hierarchical Disentanglement for Fine-Grained Object Generation and Discovery
FineGAN: Unsupervised Hierarchical Disentanglement for Fine-Grained Object Generation and Discovery
复制标题
DOI：
10.1109/cvpr.2019.00665
复制
发表时间：
2018-11
期刊：
影响因子：
--
通讯作者：
Krishna Kumar Singh;Utkarsh Ojha;Yong Jae Lee
中科院分区：
文献类型：
--
作者：
Krishna Kumar Singh;Utkarsh Ojha;Yong Jae Lee
我们提出了 FineGAN，一种新颖的无监督 GAN 框架，它解开背景、物体形状和物体外观，以分层生成细粒度物体类别的图像。为了在没有监督的情况下理清因素，我们的关键思想是使用信息论将每个因素与潜在代码相关联，并以特定方式调节代码之间的关系以产生所需的层次结构。通过大量的实验，我们表明 FineGAN 实现了所需的解缠结，以生成属于细粒度类别的鸟类、狗和汽车的逼真且多样化的图像。使用 FineGAN 的自动学习功能，我们还对真实图像进行聚类，作为解决无监督细粒度对象类别发现这一新问题的首次尝试。我们的代码/模型/演示可以在 https://github.com/kkanshul/finegan 找到
We propose FineGAN, a novel unsupervised GAN framework, which disentangles the background, object shape, and object appearance to hierarchically generate images of fine-grained object categories. To disentangle the factors without supervision, our key idea is to use information theory to associate each factor to a latent code, and to condition the relationships between the codes in a specific way to induce the desired hierarchy. Through extensive experiments, we show that FineGAN achieves the desired disentanglement to generate realistic and diverse images belonging to fine-grained classes of birds, dogs, and cars. Using FineGAN's automatically learned features, we also cluster real images as a first attempt at solving the novel problem of unsupervised fine-grained object category discovery. Our code/models/demo can be found at https://github.com/kkanshul/finegan