Stein Unbiased GrAdient estimator of the Risk (SUGAR) for multiple parameter selection

Abstract : Algorithms to solve variational regularization of ill-posed inverse problems usually involve operators that depend on a collection of continuous parameters. When these operators enjoy some (local) regularity, these parameters can be selected using the so-called Stein Unbiased Risk Estimate (SURE). While this selection is usually performed by exhaustive search, we address in this work the problem of using the SURE to efficiently optimize for a collection of continuous parameters of the model. When considering non-smooth regularizers, such as the popular l1-norm corresponding to soft-thresholding mapping, the SURE is a discontinuous function of the parameters preventing the use of gradient descent optimization techniques. Instead, we focus on an approximation of the SURE based on finite differences as proposed in (Ramani et al., 2008). Under mild assumptions on the estimation mapping, we show that this approximation is a weakly differentiable function of the parameters and its weak gradient, coined the Stein Unbiased GrAdient estimator of the Risk (SUGAR), provides an asymptotically (with respect to the data dimension) unbiased estimate of the gradient of the risk. Moreover, in the particular case of soft-thresholding, it is proved to be also a consistent estimator. This gradient estimate can then be used as a basis to perform a quasi-Newton optimization. The computation of the SUGAR relies on the closed-form (weak) differentiation of the non-smooth function. We provide its expression for a large class of iterative methods including proximal splitting ones and apply our strategy to regularizations involving non-smooth convex structured penalties. Illustrations on various image restoration and matrix completion problems are given.
Liste complète des métadonnées

Cited literature [67 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-00987295
Contributor : Charles-Alban Deledalle <>
Submitted on : Wednesday, August 6, 2014 - 7:00:18 PM
Last modification on : Friday, April 19, 2019 - 1:28:12 PM
Document(s) archivé(s) le : Wednesday, November 26, 2014 - 1:02:16 AM

Files

sugar_hal.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00987295, version 2
  • ARXIV : 1405.1164

Citation

Charles-Alban Deledalle, Samuel Vaiter, Jalal M. Fadili, Gabriel Peyré. Stein Unbiased GrAdient estimator of the Risk (SUGAR) for multiple parameter selection. SIAM Journal on Imaging Sciences, Society for Industrial and Applied Mathematics, 2014, 7 (4), pp.2448-2487. ⟨hal-00987295v2⟩

Share

Metrics

Record views

1385

Files downloads

652