A computer vision prototype to detect key facial points in a human face. This is called Face Mesh Detection, it is a machine learning model for detecting & analyzing key facial features from images. It is like a 3D model of a face.
MediaPipe Face Mesh is a face geometry solution that estimates 468 3D face landmarks in real-time even on mobile devices. It employs machine learning (ML) to infer the 3D surface geometry, requiring only a single camera input without the need for a dedicated depth sensor. Utilizing lightweight model architectures together with GPU acceleration throughout the pipeline, the solution delivers real-time performance critical for live experiences.
Additionally, the solution is bundled with the Face Geometry module that bridges the gap between the face landmark estimation and useful real-time augmented reality (AR) applications. It establishes a metric 3D space and uses the face landmark screen positions to estimate face geometry within that space.
The face geometry data consists of common 3D geometry primitives, including a face pose transformation matrix and a triangular face mesh. Under the hood, a lightweight statistical analysis method called Procrustes Analysis is employed to drive a robust, performant, and portable logic. The analysis runs on a CPU and has a minimal speed/memory footprint on top of the ML model inference.
- Detect human face Mesh.
- Able to analyze facial key points.
- Able to build a geometry-based face recognition application with this prototype.
- Python 3.8
- OpenCV
- CVzone
CVzone is the main Framework that is used in this prototype, it is a computer vision package that makes us easy to run like face detection, hand tracking, pose estimation, etc., and also image processing and other AI functions.
Python is the main programming language of this prototype, and OpenCV and CVzone are the major Frameworks.
-
Installing OpenCV
pip install opencv-python
-
Installing cvzone using GitHub, (If "pip install cvzone" does not work)
pip install git+https://github.com/cvzone/cvzone.git
