This is the official (Pytorch) implementation for the paper "Synthetic-to-Real Camouflaged Object Detection", ACM MM 2025.

The main python libraries we use:

• Python 3.10.11
• torch 2.0.1
• numpy 1.24.3

Please create a directory named Dataset in current directory, then download the following datasets and unzip into Dataset:

Source (Synthetic):

• HKU-IS ( GoogleDrive | BaiduYun )
• CAMO + NC4K + CHAMELEON (CNC) ( GoogleDrive | BaiduYun )

We generate synthetic COD images using the LAKE-RED project. You can also use it to generate your own synthetic COD data.

Target (Real):

• COD10K-train ( GoogleDrive | BaiduYun )

Test (Real):

• COD10K-test ( GoogleDrive | BaiduYun )

Val (Real):

• CAMO ( GoogleDrive | BaiduYun )

After finishing above steps, your directory structure of code may like this:

S2R-COD/
    |-- Dataset/
        |-- Source/
            |-- CNC/
            |-- HKU-IS/
        |-- Target/
        |-- Test/
        |-- Val/
    |-- Eval/
    |-- Src/
    CLS.py
    MyTest.py
    MyTrain.py
    README.md

• Downloading ResNet weights ( GoogleDrive | BaiduYun ) pretrained on ImageNet dataset for SINet, and move it into ./Src/model/SINet. Res2Net ( GoogleDrive | BaiduYun ) for SINet-v2 and move it into ./Src/model/SINetV2.

• You can use the following command to start training:

python MyTrain.py --network [SINet, SINet-v2] --task [C2C, S2C] --save_model Your_save_path --source_root Your_source_path

Use the following command to generate prediction masks with a trained model:

python MyTest.py --network [SINet, SINet-v2] --model_path Your_checkpoint_path --test_save Your_mask_path

([x,y,z] means choices.)

We adopt the evaluation protocol from the DGNet project. Use the command below to evaluate your predicted masks:

python MyEval.py --pred_root Your_mask_path --txt_name Your_result_path

• C2C: SINet ( GoogleDrive | BaiduYun ) SINet-v2 ( GoogleDrive | BaiduYun )

• S2C: SINet ( GoogleDrive | BaiduYun ) SINet-v2 ( GoogleDrive | BaiduYun )

If you would like to cite our work, the following bibtex code may be helpful:

@inproceedings{luo2025synthetic,
  title={Synthetic-to-Real Camouflaged Object Detection},
  author={Luo, Zhihao and Lin, Luojun and Lin, Zheng},
  booktitle={Proceedings of the 33rd ACM International Conference on Multimedia},
  pages={8369--8378},
  year={2025}
}

https://drive.google.com/file/d/1lmR8Q-aE1emvBHYdrPhkFnW3vTX83WET/view?usp=sharing

This source code is released under the MIT license. View it here.