This demo consists of two OpenFaaS functions which you can deploy yourself.
This is a forked version of the work by Magnus Erik Hvass Pedersen - its a tweaked version of the faas-and-furious/inception-function which has been updated to classify images submitted as base64 image data in a POST body (rather than a submitted URL).
This is a very basic function which returns a static webpage which acts as a frontend to the hotdog-classifier function above.
Note that only Android or Desktops browsers are supported (Chrome/Firefox) as it uses the
getUserMediaweb API to capture the image to be classified from a local camera.