An automated system based on cryogenic probe station for integrated studies of semiconductor light-emitting structures and wafers in the range of 15 to 475 K

An automated system for integrated electrophysical and optical studies of semiconductor nanoheterostructures, which operates in a wide temperature range from 15 to 475 K, is designed. The setup is intended to measure the temperature and frequency admittance and electroluminescence spectra of light-emitting diode and laser chips formed on substrates of diameter up to 50.2 mm, and the distribution of parameters over the wafer. The setup includes the closed-cycle helium cryogenic station, LCR meter, and temperature controller. The characterization results of nanoheterostructures with InGaN/GaN multiple quantum wells, which are used for creating highly efficient white and blue light-emitting diodes, are presented.     

Irreversible deformation and the superplasticity of a TN-1 alloy during thermal cycling through the martensitic transformations ranges under loading

The influence of the thermal cycling conditions on the thermal-cycling creep of a TN-1 alloy and the related irreversible deformations is studied. The conditions under which an anomalous increase in the irreversible deformations begins are determined. The structural mechanism of the irreversible deformations of an equiatomic alloy is shown to be analogous to the structural mechanism of metal creep at high temperatures: it predominantly has a dislocation character. It is proposed to use the effect of anomalous increase in the deformation of materials with reversible martensitic transformations for forming parts made of these materials at low temperatures.     

A conceptualization of intended learning outcomes supporting self‐regulated learners in indicating learning paths

Intended learning outcomes (ILOs) indicate what learners will be able to achieve after they are taught. Traditionally, ILOs are expressed as plain text or unstructured documents. What if all ILOs of a specific course of study can be conceptualized through a structured diagrammatic technique? It was hypothesized that learners can benefit from this conceptualization in learning, especially in self‐regulated learning. The aim of this study was to investigate whether the ILOs represented in unstructured or structured formats can facilitate learners to identify learning paths. The results revealed that the mean ratings of all learning paths were statistically significantly higher with structured ILOs.     

改进HSR-FCN的服装图像识别分类算法研究

目前网络上的服装图像数量增长迅猛，对于大量服装图像实现智能分类的需求日益增加。将基于区域的全卷积网络（Region-Based Fully Convolutional Networks，R-FCN）引入到服装图像识别中，针对服装图像分类中网络训练时间长、形变服装图像识别率低的问题，提出一种新颖的改进框架HSR-FCN。新框架将R-FCN中的区域建议网络和HyperNet网络相融合，改变图片特征学习方式，使得HSR-FCN可以在更短的训练时间内达到更高的准确率。在模型中引入了空间转换网络，对输入服装图像和特征图进行了空间变换及对齐，加强了对多角度服装和形变服装的特征学习。实验结果表明，改进后的HSR-FCN模型有效地加强了对形变服装图像的学习，且在训练时间更短的情况下，比原来的网络模型R-FCN平均准确率提高了大约3个百分点，达到96.69%。