论文标题
建筑物分割口罩的机器学习正规化和多角化
Machine-learned Regularization and Polygonization of Building Segmentation Masks
论文作者
论文摘要
我们提出了一种基于机器学习的方法，用于自动正规化和建筑物分割掩模的多角化。以图像为输入，我们首先预测利用通用完全卷积网络（FCN）的构建分割图。然后，涉及生成的对抗网络（GAN），以执行建筑边界的正则化，以使其更现实，即，如果需要，则具有更多的直线大纲，这些概述构建了正确的角度。这是通过歧视器之间的相互作用来实现的，该歧视器提供了输入图像的概率和生成器，从歧视者的响应中学习以创建更真实的图像。最后，我们训练骨干卷积神经网络（CNN），该网络（CNN）适应于预测与正规建筑物分割结果中建立角落相对应的稀疏结果。在三个建筑物分割数据集上进行的实验表明，所提出的方法不仅能够获得准确的结果，而且还可以产生视觉上令人愉悦的构建大纲参数为多边形。
We propose a machine learning based approach for automatic regularization and polygonization of building segmentation masks. Taking an image as input, we first predict building segmentation maps exploiting generic fully convolutional network (FCN). A generative adversarial network (GAN) is then involved to perform a regularization of building boundaries to make them more realistic, i.e., having more rectilinear outlines which construct right angles if required. This is achieved through the interplay between the discriminator which gives a probability of input image being true and generator that learns from discriminator's response to create more realistic images. Finally, we train the backbone convolutional neural network (CNN) which is adapted to predict sparse outcomes corresponding to building corners out of regularized building segmentation results. Experiments on three building segmentation datasets demonstrate that the proposed method is not only capable of obtaining accurate results, but also of producing visually pleasing building outlines parameterized as polygons.
