On Non-Linear operators for Geometric Deep Learning - Archive ouverte HAL Access content directly
Conference Papers Year : 2022

On Non-Linear operators for Geometric Deep Learning


This work studies operators mapping vector and scalar fields defined over a manifold $\mathcal{M}$, and which commute with its group of diffeomorphisms $\text{Diff}(\mathcal{M})$. We prove that in the case of scalar fields $L^p_\omega(\mathcal{M,\mathbb{R}})$, those operators correspond to point-wise non-linearities, recovering and extending known results on $\mathbb{R}^d$. In the context of Neural Networks defined over $\mathcal{M}$, it indicates that point-wise non-linear operators are the only universal family that commutes with any group of symmetries, and justifies their systematic use in combination with dedicated linear operators commuting with specific symmetries. In the case of vector fields $L^p_\omega(\mathcal{M},T\mathcal{M})$, we show that those operators are solely the scalar multiplication. It indicates that $\text{Diff}(\mathcal{M})$ is too rich and that there is no universal class of non-linear operators to motivate the design of Neural Networks over the symmetries of $\mathcal{M}$.
Fichier principal
Vignette du fichier
main.pdf (248.89 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03711864 , version 1 (05-07-2022)
hal-03711864 , version 2 (08-02-2023)



Grégoire Sergeant-Perthuis, Jakob Maier, Joan Bruna, Edouard Oyallon. On Non-Linear operators for Geometric Deep Learning. Conference on Neural Information Processing Systems (Neurips), Dec 2022, New Orleans, United States. ⟨10.48550/arXiv.2207.03485⟩. ⟨hal-03711864v2⟩
218 View
332 Download



Gmail Facebook X LinkedIn More