arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.16337cs.LG
收藏

基于图神经网络的信号处理变分模型

Variational models for signal processing with Graph Neural Networks

Normandie Univ, UNICAEN, ENSICAEN, CNRS, GREYC(诺曼底大学,卡昂大学,卡昂国立高等工程师学校,国家科学研究中心,格雷克实验室)

Amitoz Azad, Julien Rabin, Abderrahim Elmoataz

摘要:

本文研究了利用神经网络对点云数据进行信号处理的方法。如今,图像处理和计算机视觉的最新技术主要基于在大型数据集上训练深度卷积神经网络。虽然使用图神经网络(GNN)处理点云也是如此,但重点主要是在标记数据集(如ShapeNet)上使用监督学习进行分类和分割等高级任务。然而,这样的数据集是稀缺的,并且根据目标应用程序构建起来耗时。在这项工作中,我们研究了使用变分模型的GNN处理信号的无监督学习图。我们的贡献是双重的。我们首先表明,一些现有的基于变分的算法的信号图可以制定为消息传递网络(MPN),GNN的一个特殊的实例,使它们在实践中的计算效率相比,标准的基于梯度的机器学习算法。其次,我们研究了前馈GNN的无监督学习,无论是通过直接优化逆问题,还是通过基于变分的MPN模型蒸馏。关键词:图形处理神经网络。总变差。变分方法消息传递网络。无监督学习

英文摘要:

This paper is devoted to signal processing on point-clouds by means of neural networks. Nowadays, state-of-the-art in image processing and computer vision is mostly based on training deep convolutional neural networks on large datasets. While it is also the case for the processing of point-clouds with Graph Neural Networks (GNN), the focus has been largely given to high-level tasks such as classification and segmentation using supervised learning on labeled datasets such as ShapeNet. Yet, such datasets are scarce and time-consuming to build depending on the target application. In this work, we investigate the use of variational models for such GNN to process signals on graphs for unsupervised learning. Our contributions are two-fold. We first show that some existing variational-based algorithms for signals on graphs can be formulated as Message Passing Networks (MPN), a particular instance of GNN, making them computationally efficient in practice when compared to standard gradient-based machine learning algorithms. Secondly, we investigate the unsupervised learning of feed-forward GNN, either by direct optimization of an inverse problem or by model distillation from variational-based MPN. Keywords:Graph Processing. Neural Network. Total Variation. Variational Methods. Message Passing Network. Unsupervised learning

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑