Stability of Unfolded Forward-Backward to Perturbations in Observed Data - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2023

Stability of Unfolded Forward-Backward to Perturbations in Observed Data

Abstract

We consider a neural network architecture to solve inverse problems, which is built by unfolding a forwardbackward algorithm. This algorithm is based on the minimization of an objective function which corresponds to a penalized least squares problem. In this context, ensuring stability is consistent with inverse problem theory since it guarantees both the continuity of the inversion method and its insensitivity to small noise. The latter is a critical property as deep neural networks have been shown to be vulnerable to adversarial perturbations. The main novelty of our work is to analyze the robustness of this inversion method with respect to a perturbation of the bias parameter of the network. In our architecture, the bias accounts for the observed data in the inverse problem. The analysis is performed by using tools of fixed point theory. Our theoretical results are illustrated by numerical simulations on a problem of signal restoration.
Fichier principal
Vignette du fichier
Eusipco2023_Abel.pdf (236.56 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04165350 , version 1 (18-07-2023)

Licence

Attribution

Identifiers

  • HAL Id : hal-04165350 , version 1

Cite

Cécile de Valle, Ezequiel Centofanti, Emilie Chouzenoux, Jean-Christophe Pesquet. Stability of Unfolded Forward-Backward to Perturbations in Observed Data. EUSIPCO 2023 - 31st European Signal Processing Conference, Sep 2023, Helsinki, Finland. ⟨hal-04165350⟩
46 View
97 Download

Share

Gmail Facebook X LinkedIn More