Blockwise SVD with error in the operator and application to blind deconvolution

TL;DR

This paper develops a minimax optimal nonlinear method for solving linear inverse problems with unknown signals and operators, utilizing blockwise SVD to improve performance in blind deconvolution scenarios.

Contribution

It introduces a new adaptive procedure based on blockwise SVD that achieves near-optimal convergence rates for inverse problems with operator errors, especially in blind deconvolution.

Findings

The proposed method attains minimax optimal rates up to logarithmic factors.

Blockwise SVD-based approach outperforms classical methods under high operator noise.

Application to blind deconvolution on the torus and sphere demonstrates practical effectiveness.

Abstract

We consider linear inverse problems in a nonparametric statistical framework. Both the signal and the operator are unknown and subject to error measurements. We establish minimax rates of convergence under squared error loss when the operator admits a blockwise singular value decomposition (blockwise SVD) and the smoothness of the signal is measured in a Sobolev sense. We construct a nonlinear procedure adapting simultaneously to the unknown smoothness of both the signal and the operator and achieving the optimal rate of convergence to within logarithmic terms. When the noise level in the operator is dominant, by taking full advantage of the blockwise SVD property, we demonstrate that the block SVD procedure overperforms classical methods based on Galerkin projection or nonlinear wavelet thresholding. We subsequently apply our abstract framework to the specific case of blind deconvolution on the torus and on the sphere.