Back to Works

Facial Landmark Detection

2021

Facial landmark detection in patient's heightmap images using Google's MediaPipe and integration with Unity3D. These landmarks are later used to perform registration between the point cloud from our XR application and the point cloud from the patient's CT data. the accuracy of the point cloud registration algorithm directly depends on the accuracy of predicted landmarks.

Repository
GitHub Repository
Platform
Windows / Linux / Unity3D
Stack
MediaPipe, Python, Unity3D, Barracuda, C#
Facial Landmark Detection Facial Landmark Detection