Skip to content

Code for GRSL 2022 paper. Fully Squeezed Multi-Scale Inference Network for Fast and Accurate Saliency Detection in Optical Remote Sensing Images.

Notifications You must be signed in to change notification settings

zxforchid/FSMINet

Repository files navigation

FSMINet (GRSL 2022)

Kunye Shen, Xiaofei Zhou, Bin Wan, Ran Shi, Jiyong Zhang, 'Fully Squeezed Multi-Scale Inference Network for Fast and Accurate Saliency Detection in Optical Remote Sensing Images'.

Required libraries

Python 3.7
numpy 1.18.1
scikit-image 0.17.2
PyTorch 1.4.0
torchvision 0.5.0
glob

The SSIM loss is adapted from pytorch-ssim.

Usage

  1. Clone this repo
https://github.com/Kunye-Shen/FSMINet.git
  1. We provide the predicted saliency maps (GoogleDrive or baidu extraction code: 12so.). You can download directly through the above methods, or contact us through the following email.

Architecture

FSM Module

FSM Module architecture

FSMINet

FSMINet architecture

Quantitative Comparison

Quantitative Comparison

Qualitative Comparison

Qualitative Comparison

Citation

@article{shen2022fully,
  title={Fully Squeezed Multi-Scale Inference Network for Fast and Accurate Saliency Detection in Optical Remote Sensing Images},
  author={Shen, Kunye and Zhou, Xiaofei and Wan, Bin and Shi, Ran and Zhang, Jiyong},
  journal={IEEE Geoscience and Remote Sensing Letters},
  year={2022},
  publisher={IEEE}
}

About

Code for GRSL 2022 paper. Fully Squeezed Multi-Scale Inference Network for Fast and Accurate Saliency Detection in Optical Remote Sensing Images.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages