Results 11 to 20 of about 2,214 (195)
Nailfold capillaroscopy and deep learning in diabetes 甲襞毛细血管镜检测与深度学习在糖尿病中的应用
Journal of Diabetes, Volume 15, Issue 2, Page 145-151, February 2023., 2023 Highlights Nailfold capillary images were obtained by performing nailfold video capillaroscopy in 120 adult patients with and without diabetes. Over 5000 photographs were analyzed using machine‐learning approaches, and these models were clearly able to distinguish people based on diabetes status.
Reema Shah +6 more
wiley +1 more source
The Photogrammetric Record, Volume 37, Issue 180, Page 385-409, December 2022., 2022 Among all stereo matching methods End‐to‐End (E2E) learning methods show that they can achieve the lowest and most frequent minimum errors, however, their performance drastically changes across different test‐sites, which indicates poor generalisation capabilities.
Hessah Albanwan, Rongjun Qin
wiley +1 more source
Deep Learning-assisted Accurate Defect Reconstruction Using Ultrasonic Guided Waves:一种基于深度学习的超声导波缺陷重构方法 [PDF]
, 2020 Ultrasonic guided wave technology has played a significant role in the field of nondestructive testing due to its advantages of high propagation efficiency and low energy consumption.
Da, Y. +5 more
core +1 more source
Identification of DNS covert channel based on improved convolutional neural network [PDF]
, 2020 In order to effectively identify the multiple types of DNS covert channels,the implementation of different sorts of DNS covert channel software was studied,and a detection based on the improved convolutional neural network was proposed.The experimental ...
Haoliang SUN, Meng ZHANG, Peng YANG
core +1 more source
Modulation recognition algorithm based on mixed attention prototype network
Xibei Gongye Daxue Xuebao, 2022 针对极少量带标签样本条件下的通信信号调制识别难题, 提出一种基于混合注意力原型网络的调制识别算法。综合元学习和度量学习的思想, 在原型网络框架下通过特征提取模块将信号映射至一个新的特征度量空间, 并通过比较该空间内各类原型与查询信号之间的距离确定查询信号调制样式。根据通信信号IQ分量的时序特点设计了由卷积神经网络和长短时记忆网络级联的特征提取模块, 并引入卷积注意力机制提升关键特征的权重; 采用基于Episode的训练策略, 使算法可泛化到新的信号识别任务中。仿真结果表明 ...
PANG Yiqiong +4 more
doaj +1 more source
Applying deep learning to right whale photo identification
Conservation Biology, Volume 33, Issue 3, Page 676-684, June 2019., 2019 Abstract Photo identification is an important tool for estimating abundance and monitoring population trends over time. However, manually matching photographs to known individuals is time‐consuming. Motivated by recent developments in image recognition, we hosted a data science challenge on the crowdsourcing platform Kaggle to automate the ...
Robert Bogucki +5 more
wiley +1 more source
Developmental Medicine &Child Neurology, Volume 61, Issue 3, Page E1-E35, March 2019., 2019 目的 本国际临床指南由欧洲残疾儿童学会（the European Academy of Childhood Disability，EACD）牵头制定，旨在解决发育性协调障碍（developmental coordination disorder，DCD）的定义、诊断、评估、干预以及与社会心理方面的临床应用关键问题。 方法 本指南针对五个领域的关键问题，通过文献综述和专家团队的正式讨论达成共识。为保证指南的循证基础，以“机制”、“评估”和“干预”为检索词, 对2012年更新以来提出的最新建议以及新增的“社会心理问题”和“青少年/成人”为检索词进行检索。根据牛津大学循证医学中心证据等级 (证据水平 [level of evidence, LOE]1–4) 将结果进行分类，最终转化为指南建议。并由国际 ...
Jing Hua +6 more
wiley +1 more source
智能科学与技术学报, 2021 针对皮肤镜图像数量不充足以及各类疾病之间影像数据不平衡的问题，提出一种融合类加权交叉熵损失函数和分层卷积神经网络的皮肤镜图像分类方法。首先对皮肤镜图像进行色彩恒常化处理，消除环境光源噪声；然后构建基于ResNet 50的分层卷积神经网络，并在迁移学习的基础上分别构建二分类和多分类卷积神经网络模型，根据皮肤镜图像的数量特点设置类加权交叉熵损失函数。实验结果表明，该方法具有较好的分类效果，分类准确率达到了85.94%，与未改进的分类模型ResNet 50相比，测试准确率提高了5.752%。
邵虹, 张鸣坤, 崔文成
doaj
Multiscale Global Adaptive Attention Graph Neural Network [PDF]
, 2023 Dynamic multiscale graph neural networks have high motion prediction errors due to the low correlation between the internal joints of body parts and the limited perceptual fields.
GOU Ruru, YANG Wenzhu, LUO Zifei, YUAN Yunfeng
core +1 more source
Gaoya dianqi, 2022 断路器正常分、合闸动作直接影响电力系统控制的可靠性，储能环节一旦出现问题将直接影响断路器正常工作。提出了一种基于深度学习的断路器振动信号辨识储能过程故障类型新方法。首先，将加速传感器采集到的时域振动信号进行数据扩充，再将扩充后振动信号二次采样作为训练样本，采用改进后的卷积神经网络（convolutional neural network,CNN）与长短时记忆网络（long-short term memory,LSTM）并行结构，将卷积神经网络的第1层大卷积核和多层小卷积核，均进行批量归一化（batch ...
赵书涛 +4 more
doaj