abandonsea/corrnet

[TGRS2022] [CorrNet] Lightweight Salient Object Detection in Optical Remote Sensing Images via Feature Correlation

CorrNet

This project provides the code and results for 'Lightweight Salient Object Detection in Optical Remote Sensing Images via Feature Correlation', IEEE TGRS, vol. 60, pp. 1-12, 2022. IEEE link and arxiv link

Network Architecture

Accuracy v.s. Parameters

Requirements

python 2.7 + pytorch 0.4.0 or

python 3.7 + pytorch 1.9.0

Saliency maps

We provide saliency maps and measure results (.mat) (code: m1dm) of all compared methods (code: kftm) and our CorrNet (code: fbee) on ORSSD and EORSSD datasets.

In addition, we also provide saliency maps of our CorrNet (code: lm21) on the recently published ORSI-4199 dataset.

Training

Modify paths of VGG backbone (code: ego5) in /model/vgg.py and datasets, then run train_CorrNet.py.

Pre-trained model and testing

Download the following pre-trained model, and modify paths of pre-trained model and datasets, then run test_CorrNet.py.

We also uploaded these pre-trained models in /models.

ORSSD (code: vqi7)

EORSSD (code: q5mr)

ORSI-4199 (code: va3b)

Evaluation Tool

You can use the evaluation tool (MATLAB version) to evaluate the above saliency maps.

ORSI-SOD_Summary

Citation

    @ARTICLE{Li_2022_CorrNet,
            author = {Gongyang Li and Zhi Liu and Zhen Bai and Weisi Lin and Haibin Ling},
            title = {Lightweight Salient Object Detection in Optical Remote Sensing Images via Feature Correlation},
            journal = {IEEE Transactions on Geoscience and Remote Sensing},
            volume = {60},
            pages = {1-12},
            year = {2022},
            }

If you encounter any problems with the code, want to report bugs, etc.

Please contact me at [email protected] or [email protected].