OpenCV Model Zoo – Contribution Guidelines

Welcome to the OpenCV Model Zoo! A zoo for models tuned for OpenCV DNN.

🔗 How to Contribute

1. Create a pull request on https://huggingface.co/opencv/opencv_contribution.
2. Add your files in the pull request branch following the structure below.
3. Push your changes.
4. We’ll review your code and model data in the Hugging Face Discussions and merge it.

📁 Folder Structure

Each model directory must follow this structure:

<root-folder>/                        # i.e. opencv_contribution
├── CMakeLists.txt                     # For C++ example build integration
├── LICENSE                            # Your model’s license
├── README.md                          # Model description & usage
├── <model_name>_<year><mon>.onnx      # ONNX (or supported) model file;
│                                       # use 4-digit year + three-letter month,
│                                       # e.g. edge_detection_dexined_2024sep.onnx
├── demo.cpp                           # (Optional) C++ demo
├── demo.py                            # Python demo
├── <helper_script>.py                 # (Optional) model‐loading wrapper
└── example_outputs/                   # Sample inputs & outputs
    ├── input_image.png                # At least one representative input
    ├── output_image.png               # Your model’s result on that input
    └── result.gif                     # (Optional) output animation

License

OpenCV Zoo is licensed under the Apache 2.0 license. Please refer to licenses of different models.