3D Facial Surface Geometry Estimation and Face Replacement.
- This application use the pre-trained tflite model of Google Mediapipe.
- But this app directly call the TensorFlow Lite C++ api instead of Mediapipe framework.
By default, this app uses a UVC camera for the input stream. If you want to use a recorded video file instead of a live camera, follow these steps:
$ sudo apt install libavcodec-dev libavdevice-dev libavfilter-dev libavformat-dev libavresample-dev libavutil-dev
$ make clean; make -j4 ENABLE_VDEC=true
$ ./gl2facemesh -e -v assets/sample_video.mp4