Shunyaface Library
Updates : Project documentation(beta) is on https://shunyaos.github.io/shunyaface/
ImportantNote: Following instructions are old and deprecated, it will be updated before 20th October 2020.
Restrictions on usage is relaxed it will work for few months for you . When you are sure and want to use it in a product .Please raise an issue .
Shunyaface is a face Detection/Recognition Library which simplifies face detection/recognition on devices which support Shunya OS.
It is written in C++. This library was created after studying the best features of MTCNN,Dlib,Opencv.
Some layers of MTCNN were replaced with new techniques to get faster results. To make it faster ,it uses a custom OS -Shunya which is optimised for AI libraries .
Table of contents
General info
The project aims to achieve fastest detection and recognition speed with low power hardware. Shunyaface is written for embedded-devices like RPI-4,RPI-3,Hikey970,etc and uses Shunyaos. Currently the library achieves a detection FPS of 16-17 FPS on the RaspberryPi-4.
This Library is a trial edition and will support detection/recognition for atleast 3 months (it uses a different method to determine if user has had enough trials of the code)
Screenshots
Features
- Face Detection
- Face Recognition
- Database Handling for faces
Wiki
If you want to try it yourself go here
Status
Project is: in progress. We will be constantly updating the project so as to improve speed as well as accuracy of face recognition and detection.
Contact
Created by @shunyaos. If you are insterested in buying or for any support or queries feel free to email me at [email protected]