TODO
- Numpy:
$pip3 install numpy
- PyTorch:
$pip3 install torch torchvision
- OpenCV (only needed by the test script):
$pip3 install opencv-python
- ibug.face_detection (only needed by the test script). See this repository for details: https://github.com/hhj1897/face_detection.
- ibug.face_alignment. See this repository for details: https://github.com/hhj1897/face_alignment
git clone https://github.com/ibug-group/au_detection.git
cd au_detection
pip install -e .
- To test on live video:
python au_detection_test.py [-i webcam_index]
- To test on a video file:
python au_detection_test.py [-i input_file] [-o output_file]
TODO
[1] Toisoul, Antoine, Jean Kossaifi, Adrian Bulat, Georgios Tzimiropoulos, and Maja Pantic. "Estimation of continuous valence and arousal levels from faces in naturalistic conditions." Nature Machine Intelligence 3, no. 1 (2021): 42-50.