检索范围:
排序: 展示方式:
Multiobjective image recognition algorithm in the fully automatic die bonder
JIANG Kai, CHEN Hai-xia, YUAN Sen-miao
《机械工程前沿(英文)》 2006年 第1卷 第3期 页码 313-316 doi: 10.1007/s11465-006-0026-y
关键词: clustering different recognition algorithm Algorithm multiobjective
Online recognition of drainage type based on UV-vis spectra and derivative neural network algorithm
《环境科学与工程前沿(英文)》 2021年 第15卷 第6期 doi: 10.1007/s11783-021-1430-6
• UV-vis absorption analyzer was applied in drainage type online recognition.
关键词: Drainage online recognition UV-vis spectra Derivative spectrum Convolutional neural network
丁晓青,付强
《中国工程科学》 2009年 第11卷 第10期 页码 19-24
提出一种适用于超多类手写汉字识别的新改型Adaboost算法,采用基于描述性模型的多类分类器(modified quadratic discriminant function,MQDF)作为Adaboost基元分类器,可直接进行多类分类,无需将多类问题转化为多个两类问题处理,其训练复杂度大大低于已有的多类Adaboost算法。算法提出根据广义置信度更新样本权重,实验证明这种算法适用于大规模多类分类问题。为了降低算法的识别复杂度,提出从所有训练后得到的Adaboost基元分类器组中选择一个最优的基元分类器作为最终分类器的方法进行删减。在HCL2000及THOCR-HCD数据集上进行实验证明,所提改型Adaboost算法提高了识别率的有效性,该算法的相对错误率比现有最优算法分别下降了14.3 %,8.1 %和19.5 %。
关键词: 多类Adaboost算法 手写汉字识别 广义置信度 改进的二次鉴别函数
UsingKinect for real-time emotion recognition via facial expressions
Qi-rong MAO,Xin-yu PAN,Yong-zhao ZHAN,Xiang-jun SHEN
《信息与电子工程前沿(英文)》 2015年 第16卷 第4期 页码 272-282 doi: 10.1631/FITEE.1400209
关键词: Kinect Emotion recognition Facial expression Real-time classification Fusion algorithm Support vector machine (SVM)
李德军,吕艳华,王润田
《中国工程科学》 2007年 第9卷 第5期 页码 81-84
泥浆在建筑工程中使用非常普遍,合理地控制泥浆的物理性能对于建筑工程施工及其质量控制非常 重要,通过声学方法可以有效地监测泥浆的体积浓度等物理参数。在通过声衰减和声速等介质的声学参数反演 泥浆浓度的过程中,数据拟合的好坏直接影响到反演的精确程度。通过模式识别技术,利用聚类算法,对数据 进行分类、归类处理,能有效的地提高反演的准确度。
Research on constraint-based virtual assembly technologies
YANG Rundang, WU Dianliang, FAN Xiumin, YAN Juanqi
《机械工程前沿(英文)》 2007年 第2卷 第2期 页码 243-249 doi: 10.1007/s11465-007-0043-5
关键词: movement DOF recognition algorithm assembly orientation combination
基于语音识别的电磁调控智能超表面 Article
柏林, 刘元可, 徐亮, 张政, 王强, 蒋卫祥, 仇成伟, 崔铁军
《工程(英文)》 2023年 第22卷 第3期 页码 185-190 doi: 10.1016/j.eng.2022.06.026
本研究提出并实现了一种基于人类语音识别的智能超表面平台,用于对电磁波束进行可编程调控。该智能超表面平台由数字编码超表面、语音识别模块、单片机和数模转换器(DAC)电路组成,可根据预先存储的语音指令对电磁波进行智能控制。所构建的数字编码超表面包含6 × 6 个超级子单元,每个超级子单元由4 × 4 个嵌入了变容二极管的有源数字单元组成。语音识别模块配合DAC和单片机对语音指令进行识别,并生成对应的电压序列来控制超表面。此外,在超表面的设计过程中引入遗传算法,可有效优化超表面相位分布。为了验证智能超表面平台的性能,实验展示了雷达散射截面积缩减、涡旋波束生成和波束分裂三种典型功能。所提出的方案为调控电磁波提供了一种新途径,并在电磁和声学通信之间架起了一座桥梁。
Advances in tissue state recognition in spinal surgery: a review
Hao Qu, Yu Zhao
《医学前沿(英文)》 2021年 第15卷 第4期 页码 575-584 doi: 10.1007/s11684-020-0816-3
关键词: spinal surgery tissue state recognition image force sensing bioelectrical impedance
View-invariant human action recognition via robust locally adaptive multi-view learning
Jia-geng FENG,Jun XIAO
《信息与电子工程前沿(英文)》 2015年 第16卷 第11期 页码 917-920 doi: 10.1631/FITEE.1500080
关键词: View-invariant Action recognition Multi-view learning L1-norm Local learning
Face recognition based on subset selection via metric learning on manifold
Hong SHAO,Shuang CHEN,Jie-yi ZHAO,Wen-cheng CUI,Tian-shu YU
《信息与电子工程前沿(英文)》 2015年 第16卷 第12期 页码 1046-1058 doi: 10.1631/FITEE.1500085
关键词: Face recognition Sparse representation Manifold structure Metric learning Subset selection
Named entity recognition for Chinese construction documents based on conditional random field
《工程管理前沿(英文)》 2023年 第10卷 第2期 页码 237-249 doi: 10.1007/s42524-021-0179-8
Xuhong Zhang, Haimiao Li, Xin Zhang, Meng An, Weiwei Fang, Haitao Yu
《化学科学与工程前沿(英文)》 2017年 第11卷 第2期 页码 231-237 doi: 10.1007/s11705-017-1633-3
关键词: gelator Schiff base chiral recognition gel formation gel collapse
IDEEA activity monitor: validity of activity recognition for lying, reclining, sitting and standing
null
《医学前沿(英文)》 2013年 第7卷 第1期 页码 126-131 doi: 10.1007/s11684-012-0236-0
Recent evidence demonstrates the independent negative effects of sedentary behavior on health, but there are few objective measures of sedentary behavior. Most instruments measure physical activity and are not validated as measures of sedentary behavior. The purpose of this study was to evaluate the validity of the IDEEA system’s measures of sedentary and low-intensity physical activities: lying, reclining, sitting and standing. Thirty subjects, 14 men and 16 women, aged 23 to 77 years, body mass index (BMI) between 18 to 34 kg/m2, participated in the study. IDEEA measures were compared to direct observation for 27 activities: 10 lying in bed, 3 lying on a sofa, 1 reclining in a lawn chair, 10 sitting and 3 standing. Two measures are reported, the percentage of activities accurately identified and the percentage of monitored time that was accurately labeled by the IDEEA system for all subjects. A total of 91.6% of all observed activities were accurately identified and 92.4% of the total monitored time was accurately labeled. The IDEEA system did not accurately differentiate between lying and reclining so the two activities were combined for calculating accuracy. Using this approach the IDEEA system accurately identified 96% of sitting activities for a total of 97% of the monitored sitting time, 99% and 99% for standing, 87% and 88% for lying in bed, 87% and 88% for lying on the sofa, and 83% and 83% for reclining on a lawn chair. We conclude that the IDEEA system accurately recognizes sitting and standing positions, but it is less accurate in identifying lying and reclining positions. We recommend combining the lying and reclining activities to improve accuracy. The IDEEA system enables researchers to monitor lying, reclining, sitting and standing with a reasonable level of accuracy and has the potential to advance the science of sedentary behaviors and low-intensity physical activities.
Automatically building large-scale named entity recognition corpora from Chinese Wikipedia
Jie ZHOU,Bi-cheng LI,Gang CHEN
《信息与电子工程前沿(英文)》 2015年 第16卷 第11期 页码 940-956 doi: 10.1631/FITEE.1500067
关键词: NER corpora Chinese Wikipedia Entity classification Domain adaptation Corpus selection
标题 作者 时间 类型 操作
Multiobjective image recognition algorithm in the fully automatic die bonder
JIANG Kai, CHEN Hai-xia, YUAN Sen-miao
期刊论文
Online recognition of drainage type based on UV-vis spectra and derivative neural network algorithm
期刊论文
UsingKinect for real-time emotion recognition via facial expressions
Qi-rong MAO,Xin-yu PAN,Yong-zhao ZHAN,Xiang-jun SHEN
期刊论文
Research on constraint-based virtual assembly technologies
YANG Rundang, WU Dianliang, FAN Xiumin, YAN Juanqi
期刊论文
View-invariant human action recognition via robust locally adaptive multi-view learning
Jia-geng FENG,Jun XIAO
期刊论文
Face recognition based on subset selection via metric learning on manifold
Hong SHAO,Shuang CHEN,Jie-yi ZHAO,Wen-cheng CUI,Tian-shu YU
期刊论文
Visual chiral recognition of 1,1′-binaphthol through enantioselective collapse of gel based on an amphiphilic
Xuhong Zhang, Haimiao Li, Xin Zhang, Meng An, Weiwei Fang, Haitao Yu
期刊论文
IDEEA activity monitor: validity of activity recognition for lying, reclining, sitting and standing
null
期刊论文