Iterative Thresholding Pursuit with Continuation for $\ell_{1-2}$-Regularized Sparse Recovery

2026-06-03

AI Summary

This work addresses the problem of sparse signal recovery from underdetermined and noisy linear systems by proposing an iterative thresholding pursuit algorithm with continuation (ITP-C). Built upon ℓ₁₋₂ regularization, ITP-C integrates the support identification capability of thresholding steps with second-order updates from constrained least squares, adaptively generating support sets without prior knowledge of the true sparsity level. The algorithm ensures stability and convergence through a rigorous sufficient decrease condition. Theoretical analysis establishes that ITP-C possesses a local oracle property and satisfies the Kurdyka–Łojasiewicz convergence criterion. Experimental results demonstrate that, while maintaining monotonic descent of the objective function, ITP-C significantly outperforms state-of-the-art methods in both synthetic sparse recovery and image reconstruction tasks.

Abstract

Sparse recovery aims to reconstruct sparse signals from underdetermined and possibly noisy linear measurements. Existing $\ell_{1-2}$ iterative thresholding schemes are first-order methods. We propose an iterative thresholding pursuit method with continuation (ITP-C) for $\ell_{1-2}$-regularized sparse recovery. The method goes beyond first-order thresholding by combining the active-set identification capability of the $\ell_{1-2}$ proximal step with a restricted least-squares pursuit step that provides a second-order update on the identified support. The support is generated adaptively by the thresholding update, and no prior knowledge of the true sparsity level is required. To control the possible instability of the pursuit step while preserving the descent structure of the continuation scheme, we impose a strict descent check with respect to the dynamic objective. We establish convergence of the generated sequence under the Kurdyka-Lojasiewicz framework and prove a local oracle-type property after correct support identification. Numerical experiments on synthetic sparse recovery and image reconstruction illustrate the descent preservation of the proposed safeguard and demonstrate the improved recovery performance of ITP-C over the state-of-the-art baselines.

Problem

• sparse recovery
• underdetermined measurements
• noisy measurements
• ℓ₁₋₂ regularization
• signal reconstruction

Innovation

• iterative thresholding pursuit
• ℓ₁₋₂ regularization
• second-order update
• active-set identification
• continuation method