Consultancy Needed for a quick POC

We have our Tensorflow (not Tensorflow RT) based Facial recognition model. We want to integrate this as a plugin in a DeepStream Pipeline to find a face, track it & recognize it & indicate this recognition by anyway (write to HTML, put a bounding box, whatever).

I understand we need to convert the Tensorflow to UFF format to be used. This should be part of the scope.

We need to see the performance for Jetson Nano if it can be quick enough to do the facial recognition using our model.

Anyone interested to work with us on this, please PM me with an offer, we can start tomorrow.



Here is some benchmark result of the Jetson Nano for your reference: