This is an prototype of how we can use Machine learning model with Open CV to centre face in a video stream. This can be used in online meeting to make sure that the face is in the centre of the screen. This make the attendees more comfortable.
This can be implemented into Webex, Zoom, Skype, Goole Meet, etc.
Above video Demonsrates how the centre face works. where Leftmost video is Centre Faced and Righmost is a normal View
You must be thinking that, This is similar to Apple's Centre Stage but yeah, It can be but it's different from Centre stage. As in centre stage mulitple user are recognised at the same time But in this case only one user is recognised at a time. In Centre stage camera dosen't centre you face but it centre your whole body. In Centre Face, your face is recognised and then it centres your face.
Note: This is an prototype not a real product so, there might be some errors
- To start using this, you need to install OpenCV and Mediapipe.
pip install mediapipe
pip install opencv-python
or
pip install -r requirements.txt - Run the python code to see the magic.
python3 CentreFace.py
That's all. No extra steps needed!
I welcome contributions. If you have any questions or suggestions, please open an issue, create a pull request or can you above contact information.