首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于多图学习的情感图像检索研究
引用本文:逯波,段晓东,王存睿,李泽东.基于多图学习的情感图像检索研究[J].大连民族学院学报,2016,18(5):509-512.
作者姓名:逯波  段晓东  王存睿  李泽东
作者单位:1.大连民族大学 计算机科学与工程学院,辽宁 大连 116605;
2.东北大学 系统科学研究所,沈阳 110819
基金项目:国家自然科学基金项目(61370146,61672132,61602085);辽宁省科技计划项目(2013405003)。
摘    要:构建了一个统一的多图学习框架,来验证在不同类别情感图像中,使用不同级别特征在情感图像检索上的性能表现。首先,提取每个图像在不同层级上的共有特征,其中,从元素级别提取的一般特征作为底层特征;可解释的属性特征作为中层特征;而情感图像的语义感念描述作为高层特征。其次,为每种类型的特征构建一个图模型来验证情感图像检索的性能。最后,将多个图模型合并在一个规范化的框架内来学习每个图模型的优化权重。通过在5个不同数据集上得到的实验结果验证了所提方法的有效性。

关 键 词:情感图像检索  图像情感  多图学习  

Affective Image Retrieval Based on Multi-graph Learning
LU Bo,DUAN Xiao-dong,WANG Cun-rui,LI Ze-dong.Affective Image Retrieval Based on Multi-graph Learning[J].Journal of Dalian Nationalities University,2016,18(5):509-512.
Authors:LU Bo  DUAN Xiao-dong  WANG Cun-rui  LI Ze-dong
Institution:1.School of Computer Science and Engineering, Dalian Minzu University, Dalian Liaoning 116605, China;
2.Insitute of Systems Science, Northeastern University, Shenyang 110819, China
Abstract:In this paper, we concentrate on affective image retrieval and investigate the performance of different features on different kinds of images in a multi-graph learning framework. Firstly, we extract commonly used features of different levels for each image. Generic features and features derived from elements-of-art are extracted as low-level features. Attributes and interpretable principles-of-art based features are viewed as mid-level features, while semantic concepts described by adjective noun pairs and facial expressions are extracted as high-level features. Secondly, we construct single graph for each kind of features to test the retrieval performance. Finally, we combine the multiple graphs together into a regularization framework to learn the optimized weights of each graph to efficiently explore the complementation of different features. Extensive experiments are conducted on five datasets and the results demonstrate the effectiveness of the proposed method.
Keywords:affective image retrieval  image emotion  multi-graph learning  
点击此处可从《大连民族学院学报》浏览原始摘要信息
点击此处可从《大连民族学院学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号