SggNet: A Lightweight Semantic- and Graph-Guided Network for Advanced Optical Remote Sensing Image Salient Object Detection (Under Review)

---

🚨 Notice

-> The paper presenting SggNet is currently under review. To preserve the integrity of the review process, only partial code is being shared at this stage. The full implementation, including essential model details, will be released upon acceptance. Stay tuned for updates!

---

📝 Overview

SggNet is a lightweight and efficient network for ORSI-SOD, achieving:

• Parameters: 2.70M
• FLOPs: 1.38G
• Inference Speed: 108 FPS

It demonstrates superior performance compared to state-of-the-art lightweight ORSI-SOD methods, delivering accurate saliency detection, sharper boundaries, and clearer activation maps.

---

📊 Results

Below are sample results showcasing the effectiveness of SggNet:

Example Outputs

• In Figure 1, we visualize saliency maps generated by SggNet compared to other state-of-the-art methods in challenging scenarios:

Key Metrics on EORSSD and ORSSD Datasets

Dataset	$S_m \uparrow$	$F^{max}_{\beta} \uparrow$	$F^{mean}_{\beta} \uparrow$	$F^{adp}_{\beta} \uparrow$	$E^{max}_{\phi} \uparrow$	$E^{mean}_{\phi} \uparrow$	$E^{adp}_{\phi} \uparrow$	$\mathcal{M} \downarrow$
EORSSD	0.9279	0.8770	0.8596	0.8386	0.9762	0.9689	0.9678	0.0068
ORSSD	0.9342	0.9032	0.8896	0.8884	0.9759	0.9695	0.9720	0.0111

---

📥 Installation and Usage

Clone the Repository

git clone https://github.com/LittleGrey-hjp/SggNet
cd SggNet

Install Dependencies

pip install -r requirements.txt

Training Configuration

The pretrained model(MobileNetv2) is stored in Google Drive and Baidu Drive (). After downloading, please change the file path in the corresponding code.

Run `train.sh` to train.

Testing Configuration

Our well-trained model is stored in Google Drive and Baidu Drive (). After downloading, please change the file path in the corresponding code.

Run `test.sh` to train.

Evaluation

• Evaluate SggNet: After configuring the test dataset path, run eval.sh in the srun folder for evaluation.
• PR-Curves: We provide the code for obtaining PR-Curves through detection results. Please refer to 'PR_Curve.py'.

---

📬 Contact

For questions or feedback, feel free to open an issue on GitHub or contact us via email at [email protected].

---

💡 Stay Updated

We appreciate your interest and patience! The full implementation and additional resources will be made available after the review process is complete. 🎉

---

Let me know if you’d like additional customization!