chahm/dev/: face-detection-1.0.2 metadata and description

Simple index

description_content_type text/markdown
requires_dist
  • tensorflow
  • Pillow
  • opencv-python
  • numpy
requires_python <3.13

Because this project isn't in the mirror_whitelist, no releases from root/pypi are included.

File Tox results History
face_detection-1.0.2-py3-none-any.whl
Size
466 KB
Type
Python Wheel
Python
3

face_detection

Overview:

The face_detection module is a python wrapper around Google's Blazeface model, a lightweight, deep learning face detection model tailored for mobile inference. It is also the model behind MediaPipe's face detection solution. This module serves as a replacement interface for the MediaPipe APIs, which enables us to perform custom integrations and fine-tuning of models.

Installation:

Recommended installation method is via the ML team's python package index server at https://pip.ml.prod.trulioo.com/

    pip install -i https://pip.ml.prod.trulioo.com/chahm/dev --extra-index-url https://pip.ml.prod.trulioo.com/chahm/dev --trusted-host https://pip.ml.prod.trulioo.com face_detection --upgrade

Usage

Import the module as normal and pass any images to the detect_faces method as a PIL image. The results will be in the format of a python dictionary containing any detected faces as a list of "boxes", "keypoints", and "scores"

    from face_detection import detector

    model = detector.FaceDetector()

    image = PIL.Image.open(filename)
    detect_results = model.detect_faces("my_image", image)
    print(detect_results)

Release Notes

1.0.0 5/29/2024: Initial release of face_detection module as thin wrapper around Blazeface.