Awesome

UGPL

Accepted paper in Neurips2022: Semi-Supervised Video Salient Object Detection Based on Uncertainty-Guided Pseudo Labels
Yongri Piao, Chenyang Lu, Miao Zhang, Huchuan Lu.

Prerequisites

• Ubuntu 20.04
• CUDA 11.3
• PyTorch 1.7.0
• Python 3.6

pretrained models ,code：zve9

Train/Test

Test

Modify the paths for the testing dataset and pre-trained model(10GT+50PL_best.pth).

• python test_fuse.py

Train

1.Select a certain number of ground truth, and modify the training dataset and pre-trained model(pretrain_resnet50.pth) paths to train the pseudo-label generator.

• python train.py

You can also use our pretrained model (pseudo_label.pth) to generate pseudo-labels.

2.Select a certain number of pseudo-labels, and modify the training dataset and pre-trained model paths(pretrain_resnet50.pth for RGB stream & resnet50-19c8e357.pth for OPT stream) to collaboratively train NS-GAN with the ground truth.

• python ST-train.py

Contact us

If you have any questions, please contact us (luchenyang0724@mail.dlut.edu.cn).