Solving Inverse Problems by Joint Posterior Maximization with a VAE Prior - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2019

Solving Inverse Problems by Joint Posterior Maximization with a VAE Prior

Résumé

In this paper we address the problem of solving ill-posed inverse problems in imaging where the prior is a neural generative model. Specifically we consider the decoupled case where the prior is trained once and can be reused for many different log-concave degradation models without retraining. Whereas previous MAP-based approaches to this problem lead to highly non-convex optimization algorithms, our approach computes the joint (space-latent) MAP that naturally leads to alternate optimization algorithms and to the use of a stochastic encoder to accelerate computations. The resulting technique is called JPMAP because it performs Joint Posterior Maximization using an Autoencoding Prior. We show theoretical and experimental evidence that the proposed objective function is quite close to bi-convex. Indeed it satisfies a weak bi-convexity property which is sufficient to guarantee that our optimization scheme converges to a stationary point. Experimental results also show the higher quality of the solutions obtained by our JPMAP approach with respect to other non-convex MAP approaches which more often get stuck in spurious local optima.
Fichier principal
Vignette du fichier
JPMAP-hal.pdf (855.96 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02364354 , version 1 (14-11-2019)

Licence

Paternité

Identifiants

Citer

Mario Gonzalez, Andrés Almansa, Mauricio Delbracio, Pablo Musé, Pauline Tan. Solving Inverse Problems by Joint Posterior Maximization with a VAE Prior. (KHIPU) Latin American Meeting in Artificial Intelligence, Nov 2019, Montevideo, Uruguay. ⟨hal-02364354⟩

Relations

146 Consultations
256 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More