1Xi’an Jiaotong University 2 Nankai University 3 Microsoft Research Asia
Salient object detection has been attracting a lot of interest, and recently various heuristic computational models have been designed. In this paper, we formulate saliency map computation as a regression problem. Our method, which is based on multi-level image segmentation, utilizes the supervised learning approach to map the regional feature vector to a saliency score. Saliency scores across multiple levels are finally fused to produce the saliency map. The contributions lie in two-fold. One is that we propose a discriminate regional feature integration approach for salient object detection. Compared with existing heuristic models, our proposed method is able to automatically integrate high-dimensional regional saliency features and choose discriminative ones. The other is that by investigating standard generic region properties as well as two widely studied concepts for salient object detection, i.e., regional contrast and backgroundness, our approach significantly outperforms state-of-the-art methods on six benchmark datasets. Meanwhile, we demonstrate that our method runs as fast as most existing algorithms.
- Salient Object Detection: A Discriminative Regional Feature Integration Approach, J Wang, H Jiang, Z Yuan, MM Cheng, X Hu, N Zheng, IJCV, 123(2):251–268, 2017. [pdf] [Official version] [bib]
- MATLAB Code: contains the full pipeline of our approach, including the training and testing phases.
- C++ Code: includes the testing code only. Carefully engineered code that runs 100x faster than the above Matlab code.
- Pre-trained Models: MATLAB model (~81MB), C++ Model (~91MB)
- Ground truth of MSRA-B Salient Object Dataset
Split in our paper: training set, testing set