Learning to Detect a Salient Object

被引:1270
作者
Liu, Tie [1 ,2 ]
Yuan, Zejian [3 ]
Sun, Jian [4 ]
Wang, Jingdong [5 ]
Zheng, Nanning [3 ]
Tang, Xiaoou [6 ]
Shum, Heung-Yeung [7 ]
机构
[1] Xi An Jiao Tong Univ, Inst Artificial Intelligence & Robot, Beijing 100193, Peoples R China
[2] IBM Res China, Analyt & Optimizat Dept, Beijing 100193, Peoples R China
[3] Xi An Jiao Tong Univ, Inst Artificial Intelligence & Robot, Xian 710049, Peoples R China
[4] Microsoft Res Asia, Visual Comp Grp, Beijing 100190, Peoples R China
[5] Microsoft Res Asia, Media Comp Grp, Beijing 100190, Peoples R China
[6] Chinese Univ Hong Kong, Dept Informat Engn, Shatin, Hong Kong, Peoples R China
[7] Microsoft Corp, On Line Serv Div, R&D, Redmond, WA 98052 USA
基金
中国国家自然科学基金;
关键词
Salient object detection; conditional random field; visual attention; saliency map; VISUAL-ATTENTION; MODEL;
D O I
10.1109/TPAMI.2010.70
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In this paper, we study the salient object detection problem for images. We formulate this problem as a binary labeling task where we separate the salient object from the background. We propose a set of novel features, including multiscale contrast, center-surround histogram, and color spatial distribution, to describe a salient object locally, regionally, and globally. A conditional random field is learned to effectively combine these features for salient object detection. Further, we extend the proposed approach to detect a salient object from sequential images by introducing the dynamic salient features. We collected a large image database containing tens of thousands of carefully labeled images by multiple users and a video segment database, and conducted a set of experiments over them to demonstrate the effectiveness of the proposed approach.
引用
收藏
页码:353 / 367
页数:15
相关论文
共 48 条
[1]  
[Anonymous], VISUAL ATTENTION MOD
[2]  
[Anonymous], 2009, P IEEE INT C COMP VI
[3]  
[Anonymous], 2006, P SIGCHI C HUM FACT
[4]  
[Anonymous], P 2 INT WORKSH BIOL
[5]  
[Anonymous], ADV NEURAL INFORM PR
[6]  
[Anonymous], P IEEE CS C COMP VIS
[7]  
[Anonymous], 2003, P 11 ACM INT C MULTI, DOI DOI 10.1145/957013.957094
[8]  
[Anonymous], P ACM SIGGRAPH
[9]  
[Anonymous], P IEEE CS C COMP VIS
[10]  
[Anonymous], P ACM SIGGRAPH