/cvpr2019_Pyramid-Feature-Attention-Network-for-Saliency-detection

code and model of Pyramid Feature Selective Network for Saliency detection

Primary LanguageJupyter Notebook

cvpr2019_Pyramid-Feature-Attention-Network-for-Saliency-detection

Source code for our CVPR 2019 paper "Pyramid Feature Attention Network for Saliency detection" by Ting Zhao and Xiangqian Wu. (ArXiv paper link)

Pipline

Download Saliency Maps

We provide our saliency maps of benchmark datasets used in the paper for convenience.

Google: link

Baidu: link extraction：9yt5

Setup

Install dependencies:

  Tensorflow (-gpu)
  Keras
  numpy
  opencv-python
  matplotlib

Usage:

  train:
  python train.py --train_file=train_pair.txt --model_weights=model/vgg16_no_top.h5
  test:
  jupyter notebook
  run dome.ipynb

Result

quantitative table visual

If you think this work is helpful, please cite

@inproceedings{zhao2019pyramid,
    title = {Pyramid Feature Attention Network for Saliency detection},
    author={Ting Zhao and Xiangqian Wu},
    booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
    year = {2019}
}