论文标题
通过（图）卷积神经网络处理不完整的图像
Processing of incomplete images by (graph) convolutional neural networks
论文作者
论文摘要
我们研究了从不完整图像训练神经网络的问题，而无需替换缺失值。为此，我们首先将图像表示为图，其中缺少像素完全忽略了。使用空间图卷积网络（SGCN）处理图像图像表示 - 一种类型的图形卷积网络，这是对在图像上运行的经典CNN的正确概括。一方面，我们的方法避免了丢失数据插补的问题，另一方面，CNNS和SGCN之间存在自然的对应关系。实验证实，我们的方法的表现要比类比CNN更好，并且在典型的分类和重建任务上插入了缺失值。
We investigate the problem of training neural networks from incomplete images without replacing missing values. For this purpose, we first represent an image as a graph, in which missing pixels are entirely ignored. The graph image representation is processed using a spatial graph convolutional network (SGCN) -- a type of graph convolutional networks, which is a proper generalization of classical CNNs operating on images. On one hand, our approach avoids the problem of missing data imputation while, on the other hand, there is a natural correspondence between CNNs and SGCN. Experiments confirm that our approach performs better than analogical CNNs with the imputation of missing values on typical classification and reconstruction tasks.
