We are based in Gurgaon and have already built a Face Recognition solution in Linux.
We are now looking for Android Face Recognition Developer with strong experience in Deep learning (in platforms such as Tensorflow, MLKit etc.) using Android Platform. Following are the mandatory requirements (as we have achieved the same with inhouse development):
1. Solution will pull facial data from cloud (via AWS queue polling) in runtime and train/encode the system. Developer must know how to integrate with cloud APIs or Queues to pull face images and relevant data
2. Solution will identify known faces in runtime with live videos
3. Solution will identify unknown faces and mark them for future processing. Any duplicates etc. will need to be removed to clearly distinguish known from unknown
4. Solution will send the best results to cloud
5. Face Detection and recognition accuracy should be more than 99.5%
[login to view URL] duration must be less than 0.3 seconds
7. Detection of face must be realtime with no lag
8. Live video must not be laggy. We are happy to workout the best Android hardware required for the purpose.
There are additional upcoming requirements such as integration with access controls (Wiegand, relay interface etc.), temperature sensor and qr code scanning, though can be delivered as the separate gig once face recognition is delivered successfully .
Please bid only if you have built before and feel free to quote with your experience while ignoring budget.