Extreme Value Analysis of Empirical Frame Coefficients and Implications for Denoising by Soft-Thresholding

TL;DR

This paper analyzes the extreme value distribution of empirical frame coefficients in Gaussian noise to improve denoising techniques, providing theoretical foundations for thresholding methods across various frames.

Contribution

It derives the asymptotic distribution of maximum frame coefficients in noisy data, enabling universal and frame-specific thresholding strategies for signal denoising.

Findings

Gumbel law applies to many frames like wavelets and curvelets

Highly redundant frames may have different limiting laws

Results support asymptotically sharp confidence regions for denoising

Abstract

Denoising by frame thresholding is one of the most basic and efficient methods for recovering a discrete signal or image from data that are corrupted by additive Gaussian white noise. The basic idea is to select a frame of analyzing elements that separates the data in few large coefficients due to the signal and many small coefficients mainly due to the noise \epsilon_n. Removing all data coefficients being in magnitude below a certain threshold yields a reconstruction of the original signal. In order to properly balance the amount of noise to be removed and the relevant signal features to be kept, a precise understanding of the statistical properties of thresholding is important. For that purpose we derive the asymptotic distribution of max_{\omega \in \Omega_n} |<\phi_\omega^n,\epsilon_n>| for a wide class of redundant frames (\phi_\omega^n: \omega \in \Omega_n}. Based on our theoretical results we give a rationale for universal extreme value thresholding techniques yielding asymptotically sharp confidence regions and smoothness estimates corresponding to prescribed significance levels. The results cover many frames used in imaging and signal recovery applications, such as redundant wavelet systems, curvelet frames, or unions of bases. We show that `generically' a standard Gumbel law results as it is known from the case of orthonormal wavelet bases. However, for specific highly redundant frames other limiting laws may occur. We indeed verify that the translation invariant wavelet transform shows a different asymptotic behaviour.