Most compressive sensing (CS) reconstruction methods can be divided into two categories, i.e. model-based methods and classical deep network methods. By unfolding the iterative optimization algorithm for model-based methods onto networks, deep unfolding methods have the good interpretation of model-based methods and the high speed of classical deep network methods. In this article, to solve the visual image CS problem, we propose a deep unfolding model dubbed AMP-Net. Rather than learning regularization terms, it is established by unfolding the iterative denoising process of the well-known approximate message passing algorithm. Furthermore, AMP-Net integrates deblocking modules in order to eliminate the blocking artifacts that usually appear in CS of visual images. In addition, the sampling matrix is jointly trained with other network parameters to enhance the reconstruction performance. Experimental results show that the proposed AMP-Net has better reconstruction accuracy than other state-of-the-art methods with high reconstruction speed and a small number of network parameters.

In the past decade, sparse and low-rank recovery have drawn much attention in many areas such as signal/image processing, statistics, bioinformatics and machine learning. To achieve sparsity and/or low-rankness inducing, the ℓ1 norm and nuclear norm are of the most popular regularization penalties due to their convexity. While the ℓ1 and nuclear norm are convenient as the related convex optimization problems are usually tractable, it has been shown in many applications that a nonconvex penalty can yield significantly better performance. In recent, nonconvex regularization based sparse and low-rank recovery is of considerable interest and it in fact is a main driver of the recent progress in nonconvex and nonsmooth optimization. This paper gives an overview of this topic in various fields in signal processing, statistics and machine learning, including compressive sensing (CS), sparse regression and variable selection, sparse signals separation, sparse principal component analysis (PCA), large covariance and inverse covariance matrices estimation, matrix completion, and robust PCA. We present recent developments of nonconvex regularization based sparse and low-rank recovery in these fields, addressing the issues of penalty selection, applications and the convergence of nonconvex algorithms. Code is available at https://github.com/FWen/ncreg.git.

Blind image deblurring is a long standing challenging problem in image processing and low-level vision. Recently, sophisticated priors such as dark channel prior, extreme channel prior, and local maximum gradient prior, have shown promising effectiveness. However, these methods are computationally expensive. Meanwhile, since these priors involved subproblems cannot be solved explicitly, approximate solution is commonly used, which limits the best exploitation of their capability. To address these problems, this work firstly proposes a simplified sparsity prior of local minimal pixels, namely patch-wise minimal pixels (PMP). The PMP of clear images is much more sparse than that of blurred ones, and hence is very effective in discriminating between clear and blurred images. Then, a novel algorithm is designed to efficiently exploit the sparsity of PMP in deblurring. The new algorithm flexibly imposes sparsity inducing on the PMP under the maximum a posterior (MAP) framework rather than directly uses the half quadratic splitting algorithm. By this, it avoids non-rigorous approximation solution in existing algorithms, while being much more computationally efficient. Extensive experiments demonstrate that the proposed algorithm can achieve better practical stability compared with state-of-the-arts. In terms of deblurring quality, robustness and computational efficiency, the new algorithm is superior to state-of-the-arts. Code for reproducing the results of the new method is available at https://github.com/FWen/deblur-pmp.git.

This paper addresses the robust reconstruction problem of a sparse signal from compressed measurements. We propose a robust formulation for sparse reconstruction that employs the 11-norm as the loss function for the residual error and utilizes a generalized nonconvex penalty for sparsity inducing. The 11-loss is less sensitive to outliers in the measurements than the popular 12-loss, while the nonconvex penalty has the capability of ameliorating the bias problem of the popular convex LASSO penalty and thus can yield more accurate recovery. To solve this nonconvex and nonsmooth minimization formulation efficiently, we propose a first-order algorithm based on alternating direction method of multipliers. A smoothing strategy on the 11-loss function has been used in deriving the new algorithm to make it convergent. Further, a sufficient condition for the convergence of the new algorithm has been provided for generalized nonconvex regularization. In comparison with several state-of-the-art algorithms, the new algorithm showed better performance in numerical experiments in recovering sparse signals and compressible images. The new algorithm scales well for large-scale problems, as often encountered in image processing.

Accurate localization in harsh indoor environments has long been a challenging problem due to the presence of multipath. Since joint direction-of-arrival (DOA) and time delay (TD) estimation has the capability to separate the line-of-sight signal from multipath signals in the TD space, it has recently become a key technique for accurate indoor localization in next-generation WiFi and 5G networks. This paper addresses the problem of joint azimuth, elevation, and TD estimation of multiple reflections of a known signal. First, we propose an efficient approximate maximum likelihood algorithm for this problem, which updates the DOA and TD parameters alternatingly. This algorithm applies to arbitrarily distributed (planar or 3-D) arrays. Then, we present the closed-form Cramer-Rao bound for joint DOA and TD estimation, based on which we provide further analysis to show the benefit of joint DOA and TD estimation over DOA-only estimation. Although the benefit of joint estimation has been empirically shown long ago, our analysis is the first theoretical proof of it. Finally, simulation results have been provided to demonstrate the theoretical finding and the effectiveness of the new algorithm. Matlab code for the new algorithm is available at https://github.com/FWen/JADE.git.

Consistent and accurate global positioning is a crucial problem for autonomous vehicles and robots. It is especially challenging in situations with global navigation satellite system (GNSS) being intermittently degraded. We propose an adaptive fusion system, namely GVINS (GNSS/visual-inertial navigation system), which adaptively fuses GNSS and visual-inertial odometry (VIO) to achieve consistent and accurate global positioning, even in GNSS intermittently degraded scenarios. Compared with existing methods, GVINS can provide positioning under an Earth-fixed geographic coordinate rather than a local tangent plane (LTP) coordinate, as long as any GNSS measurement presents in the trajectory. To adaptively fuse VIO and GNSS, we first use an inertial measurement unit (IMU) pre-integration based depth uncertainty estimation method to evaluate the accuracy of VIO. Then, in the optimization backend, we perform an innovative over-parameterized, 15 degree-of-freedom (DoF) pose-graph fusion. An alternating minimization (AM) algorithm is used to efficiently solve this problem. Evaluation results on both public and custom-built datasets demonstrate that, GVINS outperforms state-of-the-art fusion methods in both accuracy and stability in GNSS intermittently degraded environments.