中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
机构
采集方式
内容类型
发表日期
学科主题
筛选

浏览/检索结果: 共19条，第1-10条 帮助

条数/页： 排序方式：
MLDA: Multi-Loss Domain Adaptor for Cross-Session and Cross-Emotion EEG-Based Individual Identification 期刊论文  OAI收割
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, 2023, 卷号: 27, 期号: 12, 页码: 5767-5778
作者:  Miao, Yifan;  Jiang, Wanqing;  Su, Nuo;  Shan, Jun;  Jiang, Tianzi
  |  收藏  |  浏览/下载：34/0  |  提交时间：2024/03/26
Adversarial Heterogeneous Graph Neural Network for Robust Recommendation 期刊论文  OAI收割
IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS, 2023, 页码: 12
作者:  Sang, Lei;  Xu, Min;  Qian, Shengsheng;  Wu, Xindong
  |  收藏  |  浏览/下载：14/0  |  提交时间：2023/11/17
Characterizing and Understanding HGNNs on GPUs 期刊论文  OAI收割
IEEE COMPUTER ARCHITECTURE LETTERS, 2022, 卷号: 21, 期号: 2, 页码: 69-72
作者:  Yan, Mingyu;  Zou, Mo;  Yang, Xiaocheng;  Li, Wenming;  Ye, Xiaochun
  |  收藏  |  浏览/下载：22/0  |  提交时间：2022/12/07
Adversarial Multimodal Network for Movie Story Question Answering 期刊论文  OAI收割
IEEE TRANSACTIONS ON MULTIMEDIA, 2021, 卷号: 23, 页码: 1744-1756
作者:  Yuan, Zhaoquan;  Sun, Siyuan;  Duan, Lixin;  Li, Changsheng;  Wu, Xiao
  |  收藏  |  浏览/下载：38/0  |  提交时间：2021/08/15
基于景感生态学的城市 视觉空间定量化研究 学位论文  OAI收割
北京: 中国科学院生态环境研究中心, 2019
作者:  张永霖
  |  收藏  |  浏览/下载：128/0  |  提交时间：2020/07/30
Determining the nonlocal parameters of a MI-CRO/nanostructure by the shifts of resonant frequencies 会议论文  OAI收割
23rd International Congress on Sound and Vibration, ICSV 2016, Athens, Greece, July 10, 2016 - July 14, 2016
作者:  Liu Y;  Zhang Y(张吟)
收藏  |  浏览/下载：31/0  |  提交时间：2017/05/05
Working Memory Capacity Predicts Human Emotion Evaluation: Evidence from Event-Related Potentials 会议论文  OAI收割
Yangzhou, PEOPLES R CHINA, MAR 23-25, 2013
作者:  Gan, SZ (Gan, Shuzhen);  Yang, JF (Yang, Jianfeng);  Yang, YF (Yang, Yufang)
  |  收藏  |  浏览/下载：14/0  |  提交时间：2018/10/11
Design of motion compensation mechanism of satellite remote sensing camera (EI CONFERENCE) 会议论文  OAI收割
International Symposium on Photoelectronic Detection and Imaging 2011: Space Exploration Technologies and Applications, May 24, 2011 - May 26, 2011, Beijing, China
Gu S.; Yan Y.; Xu K.; Jin G.
收藏  |  浏览/下载：47/0  |  提交时间：2013/03/25
With the development of aerospace remote sensing technology  the ground resolution of remote sensing camera enhances continuously. Since there is relative motion between camera and ground target when taking pictures  the target image recorded in recording media is moved and blurred. In order to enhance the imaging quality and resolution of the camera  the image motion had to be compensated. In order to abate the effect of image motion to image quality of space camera and improve the resolution of the camera  the compensation method of image motion to space camera is researched. First  the reason of producing drift angle and adjustment principle are analyzed in this paper. This paper introduce the composition and transmission principle of image motion compensation mechanism. Second  the system adopts 80C31 as controller of drift angle  and adopts stepping motor for actuators  and adopts absolute photoelectric encoder as the drift Angle measuring element. Then the control mathematical model of the image motion compensation mechanism are deduced  and it achieve the closed-loop control of the drift angle position. At the last  this paper analyses the transmission precision of the mechanism. Through the experiment  we measured the actual precision of the image motion compensation mechanism  and compared with the theoretical analysis. There are two major contributions in this paper. First  the traditional image motion compensation mechanism is big volume and quality heavy. This has not fit for the development trend of space camera miniaturization and lightweight. But if reduce the volume and quality of mechanism  it will bring adverse effects for the precision and stiffness of mechanism. For this problem  This paper designed a image motion compensation that have some advantages such as small size  light weight at the same time  high precision  stiffness and so on. This image motion compensation can be applicable to the small optics cameras with high resolution. Second  the traditional mechanism control need to corrected  fitting and iterative for the control formula of mechanism. Only in this way  we can get the optimal control mathematical model. This paper has high precision of the control formula derived. It can achieve the high precision control without fitting  It also simplify the difficulty of control mathematical model establishment. This paper designed the range of adjusting of image motion compensation mechanism between -5 +5. Based on choosing-5  -4  -3  -2  -1  0  +1  +2  +3  +4  +4 as the expectation value of the imaginary drift angle  we get ten groups of the fact data in adjusting drift angle measured. The test results show that the precision of the drift angle control system can be achieved in 1. It can meet the system requirements that the precision of the control system is less than 3'  and it can achieve the high-precision image motion compensation. 2011 SPIE.  
Computer simulation research on cognitive mechanism of human vision (EI CONFERENCE) 会议论文  OAI收割
International Conference on Measuring Technology and Mechatronics Automation, ICMTMA 2010, March 13, 2010 - March 14, 2010, Changsha, China
Ke H.; You W.
收藏  |  浏览/下载：37/0  |  提交时间：2013/03/25
The development of intelligent system which accorded with human cognitive mechanism will exert a profound influence on national defense  economy  education  culture and etc. A computer simulation method according with cognitive mechanism of human vision was proposed on the basis of Structural Decomposition Theory. Many basic shape feature information tables of similar object from different side can determine a kind of object uniquely. System weighted basic shape features in two different information tables by membership degree and then calculated similarity. The identifying object which had the maximum similarity with it in system was what we needed. The system used 83 pictures of 6 kinds of objects with different side for training and testing. The experiment results demonstrated  task of basic shape extraction  object recognition and others can be completed effectively with the method in paper. The method in paper can be a new attempt on computer simulation method according with cognitive mechanism of human vision from cognitive field. 2010 Crown Copyright.  
Method of auto multi-exposure for high dynamic range imaging (EI CONFERENCE) 会议论文  OAI收割
2010 International Conference on Computer, Mechatronics, Control and Electronic Engineering, CMCE 2010, August 24, 2010 - August 26, 2010, Changchun, China
Piao Y.; Xu W.
收藏  |  浏览/下载：19/0  |  提交时间：2013/03/25
This paper proposes a method for calculating the multi-exposure times fast and accurately when the scene has a high dynamic range  minimizing the pictures needed to take and improving the dynamic range of the CCD system. The proposed method first measures the camera's response function referring [10]. After that the median value of the output picture is adjusted to be in the middle range of the system output by changing the exposure time. It can be inferred from the histogram of the current picture whether it's a high dynamic range scene by calculating the number of pixels under exposed and over exposed. Once again the under and over exposed pixels are adjusted to be in the middle according to the camera's response function. Finally different exposed pictures (2 or 3) are fused together by Gaussian function and the dynamic range is compressed by y correction. Experimental results on UNIQ-UM400 analog camera and a digital acquisition system show that the algorithm works well and solve the problem of low dynamic range of CCD camera. 2010 IEEE.