imistyrain/MTCNN
A multi-task cascaded neural network for real-time face detection and facial landmark estimation.
Not currently ranked — collecting fresh signals.
star history
This repository implements MTCNN (Multi-Task Cascaded Convolutional Neural Networks) for face detection and alignment. The system uses a three-stage cascaded architecture to progressively refine face detection and predict five facial keypoints. It runs at real-time speeds (>30 FPS) on CPU and supports deployment across Windows, Ubuntu, Mac, Android, and iOS using frameworks like Caffe, TensorFlow, and NCNN.
Frequently asked
- What is imistyrain/MTCNN?
- A multi-task cascaded neural network for real-time face detection and facial landmark estimation.
- Is MTCNN open source?
- Yes — imistyrain/MTCNN is open source, released under the MIT license.
- What language is MTCNN written in?
- imistyrain/MTCNN is primarily written in Jupyter Notebook.
- How popular is MTCNN?
- imistyrain/MTCNN has 483 stars on GitHub.
- Where can I find MTCNN?
- imistyrain/MTCNN is on GitHub at https://github.com/imistyrain/MTCNN.