Geometric Deep Learning
Geometric deep learning (GDL) focuses on developing neural network architectures that can effectively process and learn from data with inherent geometric structures, such as graphs, meshes, and point clouds. Current research emphasizes the design of equivariant models, particularly graph neural networks (GNNs), which maintain consistent representations under geometric transformations like rotations and translations, and the development of efficient pooling operators to handle large datasets. These advancements are significantly impacting various fields, improving the accuracy and efficiency of tasks ranging from molecular simulations and material science to medical image analysis and computer-aided design.
Papers
On Non-Linear operators for Geometric Deep Learning
Grégoire Sergeant-Perthuis, Jakob Maier, Joan Bruna, Edouard Oyallon
White Matter Tracts are Point Clouds: Neuropsychological Score Prediction and Critical Region Localization via Geometric Deep Learning
Yuqian Chen, Fan Zhang, Chaoyi Zhang, Tengfei Xue, Leo R. Zekelman, Jianzhong He, Yang Song, Nikos Makris, Yogesh Rathi, Alexandra J. Golby, Weidong Cai, Lauren J. O'Donnell
Medical Application of Geometric Deep Learning for the Diagnosis of Glaucoma
Alexandre H. Thiery, Fabian Braeu, Tin A. Tun, Tin Aung, Michael J. A. Girard
Geometric Deep Learning to Identify the Critical 3D Structural Features of the Optic Nerve Head for Glaucoma Diagnosis
Fabian A. Braeu, Alexandre H. Thiéry, Tin A. Tun, Aiste Kadziauskiene, George Barbastathis, Tin Aung, Michaël J. A. Girard