pengpanda / topologicalsaliency

Saliency Detection Inspired by Topological Perception Theory

Saliency Detection Inspired by Topological Perception Theory

1. Introduction

• A computational model of the topological perception theory from Lin Chen;
• A statistic experiment to reveal the close relation between topological perception theory and visual attention (saliency);
• Topological saliency prior-guided saliency detection and salient object detection framework.

2. How to use

• Unzip EdgeDetection.zip, and put model file: modelBsds.mat into .\EdgeDetection\models\forest\;
• Set your paths in Demo.m
• Run Demo.m

3. Cite this work

 @article{2021Saliency,
  title={Saliency Detection Inspired by Topological Perception Theory},
  author={ Peng, Peng  and  Yang, Kai Fu  and  Luo, Fu Ya  and  Li, Yong Jie },
  journal={International Journal of Computer Vision},
  number={11},
  year={2021},
}

4. Acknowdgement

[1] Zhu W, Liang S, Wei Y, Sun J (2014) Saliency optimization from robust background detection. In: IEEE Conference on Computer Vision and Pattern Recognition, pp 2814–2821
[2] Zitnick CL, Dollár P (2014) Edge boxes: Locating object proposals from edges. In: European conference on computer vision, Springer, pp 391–405