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.

Code

First,

python pretrain.py

to pretrain the model on DUTS.

Then,

python finetune.py

to finetune the model on DAVSOD and DAVIS.

Pretrained Models

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

Saliency Map

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

Evaluation

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

Acknowledgement

Weakly-Supervised Salient Object Detection via Scribble Annotations
Shifting More Attention to Video Salient Object Detection
Paying Attention to Video Object Pattern Understanding