Caffe Implementation (v1) for
#Learning Unsupervised Video Primary Object Segmentation through Visual Attention (CVPR19)
-
Please install our modified caffe first.
-
Download the model weight from https://drive.google.com/open?id=1aO2nAaQMy-A76NjRSQDTx53AfpqyWUCt and put dvap_agos.caffemodel into the 'model' folder.
-
Then edit paths in './test/test.py' and './test/davis_bs3.txt'.
-
After you get the original results from the network, you can use './test/davis_crf.py' to process them.
-
Our results on DAVIS16, FBMS and Youtube can be find at 'seg_results.rar'.
-
Human eye fixation data for DAVIS16, Youtube and Segtrackv2 can be found at https://drive.google.com/open?id=11LAg69lnCp7TfWFYjntPm4RK4KfgL8_Y.
If you find our method useful in your research, please cite the following papers:
@InProceedings{Wang_2019_CVPR,
author = {Wang, Wenguan and Song, Hongmei and Zhao, Shuyang and Shen, Jianbing and Zhao, Sanyuan and Hoi, Steven Chu Hong and Ling, Haibin},
title = {Learning Unsupervised Video Object Segmentation through Visual Attention},
booktitle = {CVPR},
year = {2019}
}
@InProceedings{Song_2018_ECCV,
author = {Song, Hongmei and Wang, Wenguan and Zhao, Sanyuan and Shen, Jianbing and Lam, Kin-Man},
title = {Pyramid Dilated Deeper ConvLSTM for Video Salient Object Detection},
booktitle = {ECCV},
year = {2018} }