MathLee / SeaNet

[TGRS2023] [SeaNet] Lightweight Salient Object Detection in Optical Remote Sensing Images via Semantic Matching and Edge Alignment

SeaNet

This project provides the code and results for 'Lightweight Salient Object Detection in Optical Remote-Sensing Images via Semantic Matching and Edge Alignment', IEEE TGRS, 2023. IEEE link and arxiv link Homepage

Network Architecture

Requirements

python 3.7 + pytorch 1.9.0

Saliency maps

We provide saliency maps of our SeaNet on ORSSD, EORSSD, and additional ORSI-4199 datasets in './models/saliency_maps.zip'.

Training

We use data_aug.m for data augmentation.

Modify paths of datasets, then run train_SeaNet.py.

Note: our main model is under './model/SeaNet_models.py'

Pre-trained model and testing

1. We provide the pre-trained models in './models/'.

2. Modify paths of pre-trained models and datasets.

3. Run test_SeaNet.py.

Evaluation Tool

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

ORSI-SOD_Summary

Citation

    @ARTICLE{Li_2023_SeaNet,
            author = {Gongyang Li and Zhi Liu and Xinpeng Zhang and Weisi Lin},
            title = {Lightweight Salient Object Detection in Optical Remote-Sensing Images via Semantic Matching and Edge Alignment},
            journal = {IEEE Transactions on Geoscience and Remote Sensing},
            volume = {61},
            year = {2023},
            doi = {10.1109/TGRS.2023.3235717},
            }

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

Please contact me at [e-mail] or [e-mail].