论文阅读 [TPAMI-2022] Guest Editorial: Non-Euclidean Machine Learning

论文搜索(studyai.com)

搜索论文: Guest Editorial: Non-Euclidean Machine Learning

搜索论文: http://www.studyai.com/search/whole-site/?q=Guest+Editorial:+Non-Euclidean+Machine+Learning

关键字(Keywords)

机器视觉; AI与Web

社交图谱; 三维点云

摘要(Abstract)

Over the past decade, deep learning has had a revolutionary impact on a broad range of fields such as computer vision and image processing, computational photography, medical imaging and speech and language analysis and synthesis etc.

在过去的十年里，深度学习在计算机视觉和图像处理、计算摄影、医学成像、语音和语言分析与合成等广泛领域产生了革命性的影响。.

Deep learning technologies are estimated to have added billions in business value, created new markets, and transformed entire industrial segments.

据估计，深度学习技术增加了数十亿的商业价值，创造了新市场，并改变了整个工业领域。.

Most of today’s successful deep learning methods such as Convolutional Neural Networks (CNNs) rely on classical signal processing models that limit their applicability to data with underlying Euclidean grid-like structure, e.g., images or acoustic signals.

如今，卷积神经网络（CNN）等大多数成功的深度学习方法都依赖于经典的信号处理模型，这些模型限制了它们对具有潜在欧几里德网格结构的数据的适用性，例如图像或声音信号。.

Yet, many applications deal with non-Euclidean (graph- or manifold-structured) data.

然而，许多应用程序处理非欧几里德（图形或流形结构）数据。.

For example, in social network analysis the users and their attributes are generally modeled as signals on the vertices of graphs.

例如，在社交网络分析中，用户及其属性通常被建模为图的顶点上的信号。.

In biology protein-to-protein interactions are modeled as graphs.

在生物学中，蛋白质之间的相互作用被建模为图形。.

In computer vision & graphics 3D objects are modeled as meshes or point clouds.

在计算机视觉和图形中，3D对象被建模为网格或点云。.

Furthermore, a graph representation is a very natural way to describe interactions between objects or signals.

此外，图形表示是描述对象或信号之间交互的一种非常自然的方式。.

The classical deep learning paradigm on Euclidean domains falls short in providing appropriate tools for such kind of data.

欧几里德领域的经典深度学习范式在为此类数据提供适当工具方面存在不足。.

Until recently, the lack of deep learning models capable of correctly dealing with non-Euclidean data has been a major obstacle in these fields.

直到最近，缺乏能够正确处理非欧几里德数据的深度学习模型一直是这些领域的主要障碍。.

This special section addresses the need to bring together leading efforts in non-Euclidean deep learning across all communities.

这一特别部分讨论了将所有社区的非欧几里德深度学习的主要努力结合起来的必要性。.

From the papers that the special received twelve were selected for publication.

从特别报告员收到的论文中，有12篇被挑选出来发表。.

The selected papers can naturally fall in three distinct categories: (a) methodologies that advance machine learning on data that are represented as graphs, (b) methodologies that advance machine learning on manifold-valued data, and © applications of machine learning methodologies on non-Euclidean spaces in computer vision and medical imaging.

所选论文自然可分为三类：（a）在以图形表示的数据上推进机器学习的方法；（b）在多值数据上推进机器学习的方法；（c）机器学习方法在计算机视觉和医学成像中在非欧几里德空间上的应用。.

We briefly review the accepted papers in each of the groups…

我们简要回顾了每个小组中被接受的论文。。.

作者(Authors)

[‘Stefanos Zafeiriou’, ‘Michael Bronstein’, ‘Taco Cohen’, ‘Oriol Vinyals’, ‘Le Song’, ‘Jure Leskovec’, ‘Pietro Liò’, ‘Joan Bruna’, ‘Marco Gori’]