The code is used in our IJCAI15 paper. Any 2D images generated from light-field data could be used to compute saliency maps, such as all-focus image, depth map, and focusness maps.
We also provided pre-computed saliency maps on the LFSD dataset for comparisons.
If you use it, please cite: [1] Jun Zhang, Meng Wang, Jun Gao, Yi Wang, Xudong Zhang, Xindong Wu. Saliency Detection with a Deeper Investigation of Light Field, International Joint Conference on Artificial Intelligence (IJCAI), 2212-2218, 2015.
We also thank Zhu et al. to provide code for saliency optimization. [2] Wangjiang Zhu, Shuang Liang, Yichen Wei, and Jian Sun. Saliency Optimization from Robust Background Detection. In CVPR, 2014.
For comments or questions please contact Jun Zhang (zhangjun1126@gmail.com)