dayuml / iccv17_srm Goto Github PK

This package has the source code for the paper "A Stagewise Refinement Model for Detecting Salient Objects in Images" (ICCV17).

• The paper can be found in Baidu drive or Google drive.

• We use Caffe to train our Stagewise Refinement Model.
• Our Resnet-50 model is based on the previous work simon2016cnnmodels "Imagenet pretrained models with batch normalization".
• Saliency maps are generated without any pre- or post-processing.

Train

• Download our initialized model from Baidu drive or Google drive.
• Use the code in ./train to train the network.

Test

• Download our trained model from Baidu drive or Google drive.
• Run ./test/test.m to generate saliency maps in the ./saliency_map folder.

• The saliency maps on 10 datasets including ECSSD, PASCAL-S, SOD, SED1, SED2, MSRA, DUT-OMRON, THUR15K, HKU-IS and DUTS can be found in Baidu drive or Google drive.

If you find this work useful in your research, please consider citing:

 @inproceedings{wangiccv17,
    author={Wang, Tiantian and Borji, Ali and Zhang, Lihe and Zhang, Pingping and Lu, Huchuan},
    title={A Stagewise Refinement Model for Detecting Salient Objects in Images},
    booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
    year = {2017}
 }

[email protected]