PSCFNet: Prototype Learning and Spatial Consistent Feature Fusion for Efficient Salient Object Detection

This is the official code for the IJCNN 2024 paper: "PSCFNet: Prototype Learning and Spatial Consistent Feature Fusion for Efficient Salient Object Detection".

Requirements

• Python 3.8
• PyTorch 1.10.0
• cuda 11.4

Prepare Datasets

Download the following datasets and unzip them into GT folder

• PASCAL-S
• ECSSD
• HKU-IS
• DUT-OMRON
• DUTS

Prepare Backbone

Download the following backbone and create a directory pretrain_models to place it.

• PVT-V2-B0

Training and Testing

You can train the model as followed.

    python train.py

For testing, you have to prepare a pretrained model. You can train one by yourself.

    python test.py