Good dielectric performance and broadband dielectric polarization in Ag,Nb co-doped TiO2

Due to the demand of miniaturization and integration for ceramic capacitors in electronic components market, TiO₂-based ceramics with colossal permittivity has become a research hotspot in recent years. In this work, we report that Ag⁺/Nb⁵⁺ co-doped (Ag_1/4Nb_3/4)_xTi_1−xO₂ (ANTOx) ceramics with colossal permittivity over a wide frequency and temperature range were successfully prepared by a traditional solid–state method. Notably, compositions of ANTO0.005 and ANTO0.01 respectively exhibit both low dielectric loss (0.040 and 0.050 at 1 kHz), high dielectric permittivity (9.2 × 10³ and 1.6 × 10⁴ at 1 kHz), and good thermal stability, which satisfy the requirements for the temperature range of application of X9R and X8R ceramic capacitors, respectively. The origin of the dielectric behavior was attributed to five dielectric relaxation phenomena, i.e., localized carriers' hopping, electron–pinned defect–dipoles, interfacial polarization, and oxygen vacancies ionization and diffusion, as suggested by dielectric temperature spectra and valence state analysis via XPS; wherein, electron-pinned defect–dipoles and internal barrier layer capacitance are believed to be the main causes for the giant dielectric permittivity in ANTOx ceramics.     

电网设备线夹开裂原因

某电力公司变电站用铝合金设备线夹在运行过程发生批次开裂事故,通过宏观观察、断口分析、化学成分分析、力学性能测试、冷冻模拟试验等方法,对设备线夹的开裂原因和开裂机理进行了分析。结果表明:设备线夹焊缝存在焊接缺陷,导致焊缝强度下降;设备线夹接线管底部存在积水空间,寒冷天气下积水结冰,体积膨胀,使焊缝承受设计工况外的负载而过载开裂,造成了线夹开裂。     

基于改进相干增强扩散与纹理能量测度和高斯混合模型的导光板表面缺陷检测方法

针对液晶屏(LCD)导光板表面缺陷检测方法存在漏检率和误检率较高,对产品表面复杂渐变的纹理结构适应性差的问题,提出一种基于改进相干增强扩散(ICED)与纹理能量测度和高斯混合模型(TEM-GMM)的LCD导光板表面缺陷检测方法。首先,构建ICED模型,基于结构张量引入平均曲率流扩散(MCF)滤波,使得相干增强扩散(CED)模型对缺陷的细线状纹理有良好的边缘保持效果,并利用相干性得到缺陷纹理增强和背景纹理抑制的滤波后图像;然后,根据Laws纹理能量测度(TEM)提取图像纹理特征,将图像的背景纹理特征作为离线阶段高斯混合模型(GMM)的训练数据,使用期望最大化(EM)算法估计GMM参数;最后,计算待检测图像各像素的后验概率,并将其作为在线检测阶段缺陷像素的判断依据。实验结果表明,该检测方法在导光颗粒随机、规则两种分布的缺陷图像测试数据组上的漏检率和误检率分别为3.27%、4.32%和3.59%、4.87%。所提检测方法适用范围广,可有效检测出LCD导光板表面划痕、异物、脏污和压伤等类型的缺陷。     

结合注意力的3D卷积网络脑胶质瘤分割算法

为了提升脑胶质瘤分割精度，提出一种结合注意力机制的3D卷积神经网络算法。输入3个不同尺度的图像块，经过9个卷积层和1个分类层后得到3个不同的分类结果，将分类结果与注意力学习到的权重相乘并逐体素相加得到输出。此外该算法采用了一种混合Dice损失函数与Focal损失函数的超参数损失函数。实验表明，该算法的Dice系数在整体区域、核心区域以及增强区域分别达到了95.31%、80.12%、82.25%。与已有的一种脑胶质瘤分割算法deepmedic相比，整体区域、核心区域以及增强区域的Dice系数分别提升了3%、2%、6%。在脑胶质瘤分割方面，具有重要的临床意义。     

基于卷积神经网络的单目深度估计

针对利用深度学习方法对街道图像进行深度估计，提出采用语义分割的方法解决深度图出现边界模糊等问题；估计深度通过左右视角图生成视差图进行无监督的训练。在网络模型中添加语义分割层，采取多个空洞卷积并行的结构增加感受野，同时减少了图像下采样的次数，降低了由于下采样带来的信息损失，使得的结果更加准确。这也是在深度估计中首次与空洞卷积相结合增加准确率。通过对KITTI街道数据集进行训练，与现有结果相比，除了增加检测准确性，降低错误率之外，使得效果图中的物体更加清晰，并且在效果图中还保留了一些原模型中被忽视掉的细节信息，将原始图像更加完整的表现出来。     

基于机器视觉的运煤车车号识别技术研究

在铁路运煤装车过程中为了快速、准确地识别车号,提出一种基于机器视觉的运煤车车号识别技术。将连通区域提取与投影分割法结合,实现车号的粗定位、细分割,并对图像中的断裂字符进行二次分割,构建了基于BP神经网络的分类模型进行车号识别,提升了煤炭装车的效率和精度。     

R2D2: A scalable deep learning toolkit for medical imaging segmentation

Deep learning has gained a significant popularity in recent years thanks to its tremendous success across a wide range of relevant fields of applications, including medical image analysis domain in particular. Although convolutional neural networks (CNNs) based medical applications have been providing powerful solutions and revolutionizing medicine, efficiently training of CNNs models is a tedious and challenging task. It is a computationally intensive process taking long time and rare system resources, which represents a significant hindrance to scientific research progress. In order to address this challenge, we propose in this article, R2D2, a scalable intuitive deep learning toolkit for medical imaging semantic segmentation. To the best of our knowledge, the present work is the first that aims to tackle this issue by offering a novel distributed versions of two well-known and widely used CNN segmentation architectures [ie, fully convolutional network (FCN) and U-Net]. We introduce the design and the core building blocks of R2D2. We further present and analyze its experimental evaluation results on two different concrete medical imaging segmentation use cases. R2D2 achieves up to 17.5× and 10.4× speedup than single-node based training of U-Net and FCN, respectively, with a negligible, though still unexpected segmentation accuracy loss. R2D2 offers not only an empirical evidence and investigates in-depth the latest published works but also it facilitates and significantly reduces the effort required by researchers to quickly prototype and easily discover cutting-edge CNN configurations and architectures.     

GaN中质子辐照损伤的分子动力学模拟研究

本文利用分子动力学方法研究了GaN在质子辐照下的损伤。对不同能量（1~10 keV）初级离位原子（PKA）引起的级联碰撞进行了研究，分析了点缺陷与PKA能量的关系、点缺陷随时间的演化规律、点缺陷的空间分布及点缺陷团簇的尺寸特征。研究结果表明，点缺陷的产生与PKA能量呈线性关系，不同类型的点缺陷随时间演化规律相似，点缺陷多产生在PKA径迹旁，点缺陷团簇多为孤立的点缺陷和小团簇。     

结合深度学习与支持向量机的金属零件识别

目的 在视觉引导的工业机器人自动拾取研究中，关键技术难点之一是机器人抓取目标区域的识别问题。特别是金属零件，其表面的反光、随意摆放时相互遮挡等非结构化因素都给抓取区域的识别带来巨大的挑战。因此，本文提出一种结合深度学习和支持向量机的抓取区域识别方法。方法 分别提取抓取区域的方向梯度直方图（HOG）和局部二进制模式（LBP）特征，利用主成分分析法（PCA）对融合后的特征进行降维，以此来训练支持向量机（SVM）分类器。通过训练Mask R-CNN（regions with convolutional neural network）神经网络完成抓取区域的初步分割。然后利用SVM对Mask R-CNN识别的抓取区域进行二次分类，完成对干扰区域的剔除。最后计算掩码完成实例分割，以此达到对抓取区域的精确识别。结果 对于随机摆放的铜质金属零件，本文算法与单一的Mask R-CNN及多特征融合的SVM算法就识别准确率、错检率、漏检率3个指标进行了比较，结果表明本文算法在识别准确率上较Mask R-CNN和SVM算法分别提高了7%和25%，同时有效降低了错检率与漏检率。结论 本文算法结合了Mask R-CNN与SVM两种方法，对于反光和遮挡情况具有一定的鲁棒性，同时有效地提升了目标识别的准确率。