Capsule Network (CapsNet) is a methodology with good prospects in visual tasks, since it can keep a stronger relationship of spatial information than Convolutional Neural Networks (CNNs). However, the current Capsule Network do not provide performance as expected on several benchmark data sets with complex data and backgrounds. Inspired by the multiple capsules of Diverse Capsule Network (DCNet++) and the Spatial Group-wise Enhance (SGE) mechanism, we propose the Diverse Enhanced Capsule Network (DE-CapsNet), a hierarchical architecture which uses residual convolutional layers and the position-wise dot product to build diverse enhanced primary capsules with various scales of images for complex data. The architecture adopts the Sigmoid function in a dynamic routing algorithm to get a more uniform distribution of routing coefficients which obviously distinguishes the assignment probabilities between capsules. DE-CapsNet achieved state-of-the-art accuracy on Canadian Institute For Advanced Research (CIFAR-10) in the Capsule Network field and provided better performance than the ensemble of seven CapsNets on Fashion-Modified National Institue of Standards and Technology database (F-MNIST) while achieving a 50.3% reduction in the number of parameters.

We report the formation of confined nanometer-scale regions of porous anodic alumina from thin aluminum films. Confinement is achieved by masking a thin Al film with a sputtered $\mathrm{Si}{\mathrm{O}}_2$ layer, patterned by nanoimprint lithography of a polystyrene transfer layer. Anodization in 0.3 molar oxalic acid creates vertically aligned pores that were imaged with a combination of focused ion beam milling and scanning electron microscopy. Triplets, pairs, and single pores were observed following the anodization of isolated mask features approximately $100\mathrm{nm}$ in diameter.

Electrochemical etching of silicon carbide (SiC) material has received increasing attention in recent years, due to its simple procedure, low cost and significance in the exploration of novel optoelectronic devices. In this paper, 4H-SiC substrates were electrochemically etched at a constant current density of 392.98 mA cm ⁻² in an electrolyte made up of hydrofluoric acid and deionized water. The layering of a SiC porous layer and periodic fluctuation of the voltage were witnessed for the first time, with the layering phenomenon corresponding well to the voltage period. However, no such phenomenon was observed when the SiC substrates were anodic etched under the same conditions with magnet stirring. As a result, the periodic variation of voltage was hypothesized to be the cause of regular layering during constant current electrochemical etching. Electrochemical etching in potentiostatic mode was thus performed at different voltages. We found that the diameter of the SiC nanopores increased while the thickness of the sidewall decreased with the increasing voltage. Based on the experimental findings, a model of mass transport was proposed. The mass transport process leads to periodic changes in resistance, hence the periodic change in voltage. This successfully explained the reason for the layering. Furthermore, SiC substrates were also electrochemically etched at high and low current densities, finding the existence of a threshold current density for the occurrence of the layering. Energy dispersive x-ray spectroscopy analysis showed that the composition of the SiC porous layer remained unchanged compared to the pure SiC wafer, implying that the peeling-off of the SiC porous layer obtained by electrochemical etching can be directly adopted for use on devices requiring a SiC porous structure.