论文标题
与图形注意网络的超级像素图像分类
Superpixel Image Classification with Graph Attention Networks
论文作者
论文摘要
本文介绍了使用图神经网络（GNN）模型进行图像分类的方法。我们将输入图像转换为区域邻接图（RAGS），其中区域是超类和边缘，边缘连接相邻的超级像素。我们的实验表明，将图形卷积与自发机制相结合的图形注意力网络（GAT）优于其他GNN模型。尽管由于抹布一代期间的信息丢失，原始图像分类器的性能要比GAT更好，但我们的方法论开辟了一个有趣的研究途径，超出矩形网格图像，例如360度视野全景图广泛。当前最新方法的传统卷积内核无法处理全景，而适用的超像素算法和所得的区域邻接图自然可以在没有拓扑问题的情况下为GNN提供。
This paper presents a methodology for image classification using Graph Neural Network (GNN) models. We transform the input images into region adjacency graphs (RAGs), in which regions are superpixels and edges connect neighboring superpixels. Our experiments suggest that Graph Attention Networks (GATs), which combine graph convolutions with self-attention mechanisms, outperforms other GNN models. Although raw image classifiers perform better than GATs due to information loss during the RAG generation, our methodology opens an interesting avenue of research on deep learning beyond rectangular-gridded images, such as 360-degree field of view panoramas. Traditional convolutional kernels of current state-of-the-art methods cannot handle panoramas, whereas the adapted superpixel algorithms and the resulting region adjacency graphs can naturally feed a GNN, without topology issues.
