WSVSOD

🔥🔥🔥Code for Paper in CVPR2021, Weakly Supervised Video Salient Object Detection Wangbo Zhao, Jing Zhang, Long Li, Nick Barnes, Nian Liu, Junwei Han. 🔥🔥🔥

Dataset

We provide the proposed scribble annotations for DAVIS and DAVSOD in Google Drive and Baidu Drive(86rd). SSOD provides scribble annotations for DUTS.

First,

python pretrain.py

to pretrain the model on DUTS.

Then,

python finetune.py

to finetune the model on DAVSOD and DAVIS.

We provide the pretrained "Our" model in Baidu Drive(30ja).

Here, we provide saliency maps generated from "Our" in the paper. Baidu Drive(iv0d).

We borrow the evaluation code from DAVSOD, which is based on MATLAB.