Results 11 to 20 of about 2,052 (107)

A comparative study on deep‐learning methods for dense image matching of multi‐angle and multi‐date remote sensing stereo‐images

open access: yesThe 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

An automatic workflow for orientation of historical images with large radiometric and geometric differences

open access: yesThe Photogrammetric Record, Volume 36, Issue 174, Page 77-103, June 2021., 2021
A workflow is proposed for a completely automatic orientation of historical terrestrial urban images. For such images, automatic SfM packages often fail due to large radiometric and geometric differences causing challenges with feature extraction and reliable matching.
Ferdinand Maiwald, Hans‐Gerd Maas
wiley   +1 more source

A Bibliometric Analysis and Visualization of Fractional Order Research in China over Two Decades (2001–2020)

open access: yesJournal of Mathematics, Volume 2021, Issue 1, 2021., 2021
Fractional order research has interdisciplinary characteristics and has been widely used in the field of natural sciences. Therefore, fractional order research has become an important area of concern for scholars. This paper used 2854 literatures collected from China National Knowledge Infrastructure (CNKI) database from 2001 to 2020 as the data source
Yunfei Yang   +4 more
wiley   +1 more source

Applying deep learning to right whale photo identification

open access: yesConservation 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

International clinical practice recommendations on the definition, diagnosis, assessment, intervention, and psychosocial aspects of developmental coordination disorder – Chinese (Mandarin) translation

open access: yesDevelopmental 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

PINN-type algorithm for shock capturing of hyperbolic equations(双曲型方程激波捕捉的物理信息神经网络(PINN)算法)

open access: yesZhejiang Daxue xuebao. Lixue ban, 2023
双曲型方程的数值求解算法研究一直是偏微分方程研究的热点,其中,双曲型方程的间断捕捉是难点。受物理信息神经网络(physics-informed neural networks,PINN)启发,构造了改进的PINN算法,近似求解双曲型方程的间断问题。将坐标构造的数据集作为神经网络的输入,将PINN算法中的损失函数作为训练输出值与参考解(基于细网格的熵相容格式数据)或准确解的误差值,通过网络优化,最小化损失函数,得到最优网络参数。最后用数值算例验证了算法的可行性,数值结果表明,本文算法能捕捉激波,分辨率高 ...
郑素佩(ZHENG Supei)   +3 more
doaj   +1 more source

BP神经网络的分层优化研究及其在风电功率预测中的应用

open access: yesGaoya dianqi, 2022
为改善BP神经网络算法需要大量训练数据和预测精度有限等问题,提出了以输入层、隐含层和输出层为目标的分层优化思路。首先,利用灰色模型良好的小数据趋势辨别能力对输入层数据进行处理,以更好地提炼数据内部蕴含的数学规律,压缩神经网络所需训练数据样本数量;然后,利用遗传算法优越的全局寻优能力确定隐含层的初始权值和阈值,抑制神经网络隐含层参数无法准确获取所导致的误差较大和泛化能力弱的问题;最后,采用蚁群优化算法对输出层数据进行优化,以进一步改善神经网络模型的计算精度。以波动性较强的风电功率进行算例验证,结果表明 ...
朱显辉   +4 more
doaj  

一种基于残差网络的改进网络流量分类算法

open access: yesGuangtongxin yanjiu, 2021
基于卷积神经网络的网络流量分类算法中,为了提高分类准确度,其结构设计日趋复杂,容易出现梯度下滑甚至梯度消失,导致预测准确度不升反降。文章提出了一种基于残差网络的改进流量分类算法,引入残差网络层代替传统卷积神经网络中的卷积层和池化层,不仅缓解了传统卷积网络因层次太深导致难以训练的问题,同时与传统卷积运算相比,所提出的残差网络在训练时学习到的数据特征信息更加全面,训练后的模型也更加准确。仿真结果表明,改进后的算法比常规的神经网络算法表现更佳,分类准确度从92.05%提高到了96.18%。
陆煜斌   +5 more
doaj  

基于差分进化-人工神经网络的沉积河谷地震动放大效应预测模型

open access: yesDizhen xuebao, 2022
探讨了基于差分进化-人工神经网络构建沉积河谷地震响应代理模型的可行性。首先建立沉积河谷对地震波散射的求解方法,以半圆形、V形沉积河谷为例,以入射波条件、沉积内外介质属性、场地形状为特征参数,以沉积河谷地震动放大系数为预测目标参数,构建数据集;其次,建立沉积河谷地震动放大效应人工神经网络、差分进化-人工神经网络算法预测模型,对比两种算法计算精度和稳定性,并进行了特征参数敏感性分析。结果表明:人工神经网络能较好地预测沉积河谷地震动放大效应,使差分进化-人工神经网络预测模型的精度和稳定性显著提高 ...
Sibo Meng, Jiawei Zhao, Zhongxian Liu
doaj   +1 more source

High Performance Catalyst Design by Artificial Intelligence for Acetylene Hydrochlorination

open access: yesRare Metals, Volume 45, Issue 1, January 2026.
ABSTRACT The rapid development of artificial intelligence (AI) has brought transformative contributions to society, revolutionizing various fields through its advanced computational capabilities and data‐driven approaches. In the field of materials science, AI has become a powerful tool that significantly accelerates the discovery of new materials by ...
Yuan Zhou   +12 more
wiley   +1 more source

Home - About - Disclaimer - Privacy