We propose a Deep Boosting Framework (DBF) for real-world image denoising by integrating the deep learning technique into the boosting algorithm. The DBF replaces conventional handcrafted boosting units by elaborate convolutional neural networks, which brings notable advantages in terms of both performance and speed. We design a lightweight Dense Dilated Fusion Network (DDFN) as an embodiment of the boosting unit, which addresses the vanishing of gradients during training due to the cascading of networks while promoting the efficiency of limited parameters. The capabilities of the proposed method are first validated on several representative simulation tasks including non-blind and blind Gaussian denoising and JPEG image deblocking. We then focus on a practical scenario to tackle with the complex and challenging real-world noise. To facilitate leaning-based methods including ours, we build a new Real-world Image Denoising (RID) dataset, which contains 200 pairs of high-resolution images with diverse scene content under various shooting conditions. Moreover, we conduct comprehensive analysis on the domain shift issue for real-world denoising and propose an effective one-shot domain transfer scheme to address this issue. Comprehensive experiments on widely used benchmarks demonstrate that the proposed method significantly surpasses existing methods on the task of real-world image denoising.

Emerging deep learning-based methods have enabled great progress in automatic neuron segmentation from Electron Microscopy (EM) volumes. However, the success of existing methods is heavily reliant upon a large number of annotations that are often expensive and time-consuming to collect due to dense distributions and complex structures of neurons. If the required quantity of manual annotations for learning cannot be reached, these methods turn out to be fragile. To address this issue, in this article, we propose a two-stage, semi-supervised learning method for neuron segmentation to fully extract useful information from unlabeled data. First, we devise a proxy task to enable network pre-training by reconstructing original volumes from their perturbed counterparts. This pre-training strategy implicitly extracts meaningful information on neuron structures from unlabeled data to facilitate the next stage of learning. Second, we regularize the supervised learning process with the pixel-level prediction consistencies between unlabeled samples and their perturbed counterparts. This improves the generalizability of the learned model to adapt diverse data distributions in EM volumes, especially when the number of labels is limited. Extensive experiments on representative EM datasets demonstrate the superior performance of our reinforced consistency learning compared to supervised learning, i.e., up to 400% gain on the VOI metric with only a few available labels. This is on par with a model trained on ten times the amount of labeled data in a supervised manner. Code is available at https://github.com/weih527/SSNS-Net.

The vortex wave that carries orbital angular momentum has attracted much attention due to the fact that it can provide an extra degree of freedom for optical communication, imaging and other applications. In spite of this, the method of OAM generation at high frequency still suffers from limitations, such as chromatic aberration and low efficiency. In this paper, an azimuthally symmetric electromagnetic metasurface with wide bandwidth is designed, fabricated and experimentally demonstrated to efficiently convert a left-handed (right-handed) circularly polarized incident plane wave (with a spin angular momentum (SAM) of ћ) to a right-handed (left-handed) circularly polarized vortex wave with OAM. The design methodology based on the field equivalence principle is discussed in detail. The simulation and measurement results confirm that the proposed method provides an effective way for generating OAM-carrying vortex wave with comparative performance across a broad bandwidth.

A dual-band filter using dual-capacitively loaded substrate integrated waveguide (SIW) cavities is proposed. Two conductive posts are embedded in the SIW cavity and the capacitive effect generated by the loaded posts lowers the first two resonant frequencies. Compared with the unloaded SIW cavity of the same size, the first resonant frequency is lowered by 39% (from 4.12 to 2.51 GHz), and the second resonant frequency is lowered by 20% (from 6.63 to 5.30 GHz). The two loaded posts also have the capability to tune the two resonant frequencies independently, thus two passbands of the filter using the loaded SIW cavity can be flexibly controlled. The proposed filter is fabricated through a multilayer printed circuit board (PCB) process to experimentally validate the predicted behavior.

The phase information of the echo is essential for inverse synthetic aperture radar (ISAR) imaging. Time-modulated metasurface (TMM) generates time-varying reflection phase that can modulate the phase of echo, thus enabling jamming of ISAR imaging, which has great potential for real applications. In practice, covering the whole target with time-modulated metasurface is difficult to achieve. Consequently, the echo model of a radar target partially covered with a time-modulated metasurface is developed, which extends the model of ISAR jamming with TMM fully covering to a more general and near-realistic case in this letter. Then the echoes of each component of the target are obtained by electromagnetic (EM) simulation, and the time-varying reflection phase of the 1-bit time-modulated metasurface is obtained by experimental testing. Finally, using the above data, we validate that ISAR image of the component covered metasurface can be blurred using random time modulation sequence. Meanwhile, duplicated components can be produced in the ISAR image using periodic time modulation sequence. Experimental results of range profile jamming confirm the validity of the conclusions in this letter. The partial defocusing or duplicating of the radar target in the ISAR image can reduce the recognized possibility and enhance the jamming performance, effectively.

In this paper, a simple metasurface superstrate structure is proposed to reduce mutual coupling of both E-plane and H-plane coupled patch antennas. The proposed metasurface is consisted of periodic ring slots arranged on one side of the substrate. The metasurface can suppress the surface wave and partly change the coupling mode, which can achieve isolation enhancement of more than 40dB at resonant frequency in both E/H plane coupled patch antennas. The edge-to-edge spacing between antenna elements is 2mm ( $0.023 \lambda_0$ ). The prototypes of E-plane and H-plane arrays with metasurfaces are fabricated, and the experimental measurements and the simulation results are in good agreement.