Salient Object Detection via Low-Rank and Structured Sparse Matrix Decomposition

CASIA OpenIR > 多模态人工智能系统全国重点实验室 > 视频内容安全

	Salient Object Detection via Low-Rank and Structured Sparse Matrix Decomposition
	Houwen Peng1 ; Bing Li1 ; Rongrong Ji 2; Weiming Hu1 ; Weihua Xiong 1; Congyan Lang 3
	2013
会议名称	the Twenty-Seventh AAAI Conference on Artificial Intelligence
会议录名称	AAAI Conference on Artificial Intelligence
页码	796-802
会议日期	2013
会议地点	美国
摘要	Salient object detection provides an alternative solution to various image semantic understanding tasks such as object recognition, adaptive compression and image retrieval. Recently, low-rank matrix recovery (LR) theory has been introduced into saliency detection, and achieves impressed results. However, the existing LR-based models neglect the underlying structure of images, and inevitably degrade the associated performance. In this paper, we propose a Low-rank and Structured sparse Matrix Decomposition (LSMD) model for salient object detection. In the model, a tree-structured sparsity-inducing norm regularization is firstly introduced to provide a hierarchical description of the image structure to ensure the completeness of the extracted salient object. The similarity of saliency values within the salient object is then guaranteed by the `1-norm. Finally, high-level priors are integrated to guide the matrix decomposition and enhance the saliency detection. Experimental results on the largest public benchmark database show that our model outperforms existing LRbased approaches and other state-of-the-art methods, which verifies the effectiveness and robustness of the structure cues in our model.
关键词	Salient Object Detection
收录类别	EI
文献类型	会议论文
条目标识符	http://ir.ia.ac.cn/handle/173211/4564
专题	多模态人工智能系统全国重点实验室_视频内容安全
通讯作者	Bing Li
作者单位	1.中国科学院自动化研究所 2.厦门大学 3.北京交通大学
第一作者单位	中国科学院自动化研究所
通讯作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Houwen Peng,Bing Li,Rongrong Ji,et al. Salient Object Detection via Low-Rank and Structured Sparse Matrix Decomposition[C],2013:796-802.

条目包含的文件		下载所有文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
gh45.pdf（3090KB）	会议论文		开放获取	CC BY-NC-SA	浏览下载