/VideoMasker

Secure and convenient Client-Side Video Automatic De-Identification Program

Primary LanguagePythonGNU Affero General Public License v3.0AGPL-3.0

VideoMasker

Secure and convenient Client-Side Video Automatic De-Identification Program

  1. Open the video
  2. Press the 'Analyze' button
  3. Check the thumbnail on the right and select what to exclude from de-identification (optional)
  4. Press the 'De-Id' button
  5. It's really easy, right?

Videos will not be uploaded to the server.
It is only processed on your device.

Requirements

Developed in Python 3.11

pyside6
pytorch
torchvision
transformers
scikit-learn
opencv-python
huggingface-hub
imgbeddings
ultralytics
scipy
supervision

Getting Started

git clone https://github.com/dh031200/VideoMasker
pip install -r requirements.txt
python main.py

Screenshots/Demo

Selectively de-identify targets

Selectively

Blur

Blur

Pixelate

Pixelate

Emoji

Emoji

Contributing

Feel free to fork the repository, make changes, and create pull requests. Any contributions to improve the project are welcome!

License

AGPL-3.0 license

Contact Information

If you have any questions, please contact us.
[techniflows - Contact]

Acknowledgements

Sample Video(Happy Monica Geller Scenes | Logoless 1080p) : https://www.youtube.com/watch?v=8FZ2lDSYgvs
Ultralytics-YOLOv8 : https://github.com/ultralytics/ultralytics
Roboflow-supervision : https://github.com/roboflow/supervision
ifzhang-ByteTrack : https://github.com/ifzhang/ByteTrack