Abstract:The problem of recovering a signal $\boldsymbol x\in \mathbb{R}^n$ from a quadratic system $\{y_i=\boldsymbol x^\top\boldsymbol A_i\boldsymbol x,\ i=1,\ldots,m\}$ with full-rank matrices $\boldsymbol A_i$ frequently arises in applications such as unassigned distance geometry and sub-wavelength imaging. With i.i.d. standard Gaussian matrices $\boldsymbol A_i$, this paper addresses the high-dimensional case where $m\ll n$ by incorporating prior knowledge of $\boldsymbol x$. First, we consider a $k$-sparse $\boldsymbol x$ and introduce the thresholded Wirtinger flow (TWF) algorithm that does not require the sparsity level $k$. TWF comprises two steps: the spectral initialization that identifies a point sufficiently close to $\boldsymbol x$ (up to a sign flip) when $m=O(k^2\log n)$, and the thresholded gradient descent which, when provided a good initialization, produces a sequence linearly converging to $\boldsymbol x$ with $m=O(k\log n)$ measurements. Second, we explore the generative prior, assuming that $x$ lies in the range of an $L$-Lipschitz continuous generative model with $k$-dimensional inputs in an $\ell_2$-ball of radius $r$. With an estimate correlated with the signal, we develop the projected gradient descent (PGD) algorithm that also comprises two steps: the projected power method that provides an initial vector with $O\big(\sqrt{\frac{k \log L}{m}}\big)$ $\ell_2$-error given $m=O(k\log(Lnr))$ measurements, and the projected gradient descent that refines the $\ell_2$-error to $O(\delta)$ at a geometric rate when $m=O(k\log\frac{Lrn}{\delta^2})$. Experimental results corroborate our theoretical findings and show that: (i) our approach for the sparse case notably outperforms the existing provable algorithm sparse power factorization; (ii) leveraging the generative prior allows for precise image recovery in the MNIST dataset from a small number of quadratic measurements.

Fast sparse reconstruction: Greedy inverse scale space flows

Fast Inverse-Free Generalized Sparse Bayesian Iearning Algorithm

Geophysical Data Sparse Reconstruction Based on L0-Norm Minimization

Sparsity estimation based adaptive matching pursuit algorithm

Fast Sparse Recovery Via Non-Convex Optimization

Iterative Orthogonal Matching Pursuit and Sparse Solution

A Fast Algorithm for Sparse Reconstruction Based on Shrinkage, Subspace Optimization, and Continuation

Fast OMP for Exact Recovery and Sparse Approximation

Fast Sampling for Linear Inverse Problems of Vectors and Tensors using Multilinear Extensions

Dynamic Orthogonal Matching Pursuit for Sparse Data Reconstruction

Convergence of Projected Subgradient Method with Sparse or Low-Rank Constraints

Solving Quadratic Systems with Full-Rank Matrices Using Sparse or Generative Priors

Flow Priors for Linear Inverse Problems via Iterative Corrupted Trajectory Matching

Adaptive matching pursuit for sparse signal recovery

Greedy solution of ill-posed problems: error bounds and exact inversion

An Iteratively Reweighted Algorithm for Sparse Reconstruction of Subsurface Flow Properties from Nonlinear Dynamic Data

Fast Recursive Greedy Methods for Sparse Signal Recovery

Sparse signal recovery based on forward backward operator splitting

Samplet basis pursuit: Multiresolution scattered data approximation with sparsity constraints

The greedy side of the LASSO: New algorithms for weighted sparse recovery via loss function-based orthogonal matching pursuit

A Fast Convex Conjugated Algorithm for Sparse Recovery