Recently, I have been touring around Australia to present how to do real-time face recognition with Microsoft Cognitive Services. In the presentation I go through how to reduce cost, what are some of the environmental issues and how you can leverage infrared and depth streams to improve user experience.
Here is my video from .NET User Group Sydney recorded by SSW TV (thanks Adam and Raj!):
Here is the demo project built with C#, OpenCV and Cognitive Services:
You can use either web cam or video file for facial recognition.
Here are some of the frameworks I have played around for offline face detection:
- TensorFlow.js - https://js.tensorflow.org/
Update: You can now explore Face API with Cognitive Services Exlorer.