Updated pytorch implementation of inference stage of face detection algorithm described in
Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks.
- pytorch >= 1.0
- Pillow >= 6
git clone https://github.com/yl1991/pytorch-mtcnn.git
cd pytorch-mtcnn
pip install -e .from mtcnn import detect_faces
from PIL import Image
image = Image.open('image.jpg')
bounding_boxes, landmarks = detect_faces(image)For examples see test_on_images.ipynb.
This implementation is heavily inspired by: