Phase-field Models for Lightweight Graph Convolutional Networks - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Phase-field Models for Lightweight Graph Convolutional Networks

Hichem Sahbi

Résumé

In this paper, we design lightweight graph convolutional networks (GCNs) using a particular class of regularizers, dubbed as phase-field models (PFMs). PFMs exhibit a bi-phase behavior using a particular ultra-local term that allows training both the topology and the weight parameters of GCNs as a part of a single "end-to-end" optimization problem. Our proposed solution also relies on a reparametrization that pushes the mask of the topology towards binary values leading to effective topology selection and high generalization while implementing any targeted pruning rate. Both masks and weights share the same set of latent variables and this further enhances the generalization power of the resulting lightweight GCNs. Extensive experiments conducted on the challenging task of skeleton-based recognition show the outperformance of PFMs against other staple regularizers as well as related lightweight design methods.
Fichier principal
Vignette du fichier
camera.pdf (443.99 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04274254 , version 1 (07-11-2023)

Identifiants

Citer

Hichem Sahbi. Phase-field Models for Lightweight Graph Convolutional Networks. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) / ECV, Jun 2023, Vancouver, Canada. pp.4644-4650, ⟨10.1109/CVPRW59228.2023.00489⟩. ⟨hal-04274254⟩
7 Consultations
4 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More