Input: A video/ webcam (realtime)
Output: Face detect & recognition, find the best face in all.
- Opencv 3.3.6
- Python 3
- statistics (for compute stdev)
- matplotlib (For visualize histogram if you need)
- Train your model first:
- place your images in train folder
- call function : train_and_save_model("train/")
- Edit your input video path in main.py.
- run:
- python3 main.py (see more in demo video)
The best face consider by 4 factor:
- Accurancy
- Straight face
- Contrast
- Entropy
Copyright by LinhPhan
Contact me: fb.com/linhpph
Video demo: https://youtu.be/T2cMSd4WX9M