Matrix Manifold Neural Networks++ - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2024

Matrix Manifold Neural Networks++

Résumé

Deep neural networks (DNNs) on Riemannian manifolds have garnered increasing interest in various applied areas. For instance, DNNs on spherical and hyperbolic manifolds have been designed to solve a wide range of computer vision and nature language processing tasks. One of the key factors that contribute to the success of these networks is that spherical and hyperbolic manifolds have the rich algebraic structures of gyrogroups and gyrovector spaces. This enables principled and effective generalizations of the most successful DNNs to these manifolds. Recently, some works have shown that many concepts in the theory of gyrogroups and gyrovector spaces can also be generalized to matrix manifolds such as Symmetric Positive Definite (SPD) and Grassmann manifolds. As a result, some building blocks for SPD and Grassmann neural networks, e.g., isometric models and multinomial logistic regression (MLR) can be derived in a way that is fully analogous to their spherical and hyperbolic counterparts. Building upon these works, we design fully-connected (FC) and convolutional layers for SPD neural networks. We also develop MLR on Symmetric Positive Semi-definite (SPSD) manifolds, and propose a method for performing backpropagation with the Grassmann logarithmic map in the projector perspective. We demonstrate the effectiveness of the proposed approach in the human action recognition and node classification tasks.
Fichier non déposé

Dates et versions

hal-04552453 , version 1 (19-04-2024)

Identifiants

  • HAL Id : hal-04552453 , version 1

Citer

Xuan Son Nguyen, Shuo Yang, Aymeric Histace. Matrix Manifold Neural Networks++. 12th International Conference on Learning Representations (ICLR), May 2024, Vienna, Austria. ⟨hal-04552453⟩
37 Consultations
0 Téléchargements

Partager

Gmail Mastodon Facebook X LinkedIn More