NetPrune: A sparklines visualization for network pruning - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Visual Informatics Année : 2023

NetPrune: A sparklines visualization for network pruning

Résumé

Current deep learning approaches are cutting-edge methods for solving classification tasks. Arising transfer learning techniques allows to apply large generic model to simple tasks whereas simpler models could be used. Large models raise the major problem of their memory consumption and processor usage and lead to a prohibitive ecological footprint. In that paper, we present a novel visual analytics approach to interactively prune those networks and thus limit that issue. Our technique leverages a novel sparkline matrix visualization technique as well as a novel local metric which evaluates the discriminatory power of a filter to guide the pruning process and make it interpretable. We assess the well founded of our approach through two realistic case studies and a user study. For both of them, the interactive refinement of the model led to a significantly smaller model having similar prediction accuracy than the original one.
Fichier non déposé

Dates et versions

hal-04115344 , version 1 (02-06-2023)

Identifiants

Citer

Luc-Etienne Pommé, Romain Bourqui, Romain Giot, Jason Vallet, David Auber. NetPrune: A sparklines visualization for network pruning. Visual Informatics, 2023, ⟨10.1016/j.visinf.2023.04.001⟩. ⟨hal-04115344⟩

Collections

CNRS
20 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More