6

I want to port the Java/Kotlin implementation of Firebase ml-kit to C++, for testing purposes with PC. I figured it must be some TensorFlow Lite model and I should be able to adapt from official TensorFlow source to use it.

But where is the face detection model? The three models @ https://github.com/firebase/quickstart-android/tree/master/mlkit/app/src/main/assets appear to be classification models.

And how do I parse the result to obtain face contour and facial landmark points?

I checked a few days now but haven't seen anything like that in C++.

cosmin
  • 20,422
  • 5
  • 39
  • 57
Peacher Wu
  • 161
  • 1
  • 4
  • 1
    Check the first note that appears at this [link](https://github.com/firebase/quickstart-android/blob/master/mlkit/README.md). It seems that they started offering ML Kit's on-device APIs through a new standalone SDK. At this page there is a link to download a model for face detection. You can check parameters with [Netron](https://lutzroeder.github.io/netron/) visualization tool. Now for your other question... I do not know if the [link](https://developers.google.com/ml-kit/vision/face-detection/ios) for iOs development satisfies you... Does it?? – Farmaker Sep 01 '20 at 08:49

0 Answers0