Context-measure: Contextualizing Metric for Camouflage

Chen-Yang Wang ¹,  Gepeng Ji ¹,  Song Shao ²,  Ming-Ming Cheng ^1,2,  Deng-Ping Fan ^1,2*, 

¹ VCIP, CS, Nankai University  ² Chongqing Changan Wangjiang Industrial Group Co., Ltd. 

     

🔥 Updates

• [Dec/09/2025] Project release!

🚀 Quick Evaluation

Before running the snippet, you only need to install the following minimium dependencies.

conda create -n cmeasure python=3.10
conda activate cmeasure
pip install -r requirements.txt

Then you can use the command python quick_eval.py to start. We provide two functions

• Evaluation: Running the evaluate() function produces our context-measure (general and camouflaged) along with other saliency-era evaluation metrics.

  # input original image: img, ground-truth mask: gt, predicted mask: fm
  evaluate(img, gt, fm)  # quick evaluation

• Visualization: Running the visualize() function will generate camouflage heatmap. All outputs are saved to the vis_cd/ directory.

  # input original image: img, ground-truth mask: gt
  visualize(img, gt)  # visualize camouflage degree

🛬 Citation

@article{wang2025cmeasure,
  title={Context-measure: Contextualizing Metric for Camouflage},
  author={Wang, Chen-Yang and Ji, Gepeng and Shao, Song and Cheng, Ming-Ming and Fan, Deng-Ping},
  journal={arXiv preprint arXiv:2512.07076},
  year={2025}
}

📧 Contact

Any questions and suggestions, feel free to leave issues here or send us e-mail via [email protected] or [email protected]