Face-Landmarking | Real time face landmarking using decision trees and NN | Machine Learning library

Face-Landmarking is a C++ library typically used in Artificial Intelligence, Machine Learning, Deep Learning, OpenCV applications. Face-Landmarking has no bugs, it has no vulnerabilities, it has a Permissive License and it has low support. You can download it from GitHub.

This software is written (mostly) in C++ (with some ML parts written in Python). For video capturing and processing it uses OpenCV (OpenCV is also used for initial face detection). In this project, the face landmarking is an iterative process which updates positions of key face points each frame using simple filters and decision trees. At the same time NN autoencoders ensure that the overall shape of the face stays correct. The algorithm maps 194 points on all of the detected faces each frame. The ML models have been thought using the HELEN dataset: It's just a pet project of mine, so it still requires some work. In particular I didn't have that much time (nor will :D) to conduct extended experiments. Most of the parameters (like shape and number of filters, size of NN etc.) are just mine educated guesses.