MTCNN
pytorch
implementation of inference stage of face detection algorithm described in
Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks.
Example
How to use it
Just download the repository and then do this
from src 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
.
Requirements
- pytorch 0.2
- Pillow, numpy
Credit
This implementation is heavily inspired by: