I’d like a tutorial on how to export a bounding box detector network to mxnet, (for bonus points: convert to Tensorflow), and then get it running correctly in python!

Some related questions exist for pythonizing simple recurrent and classification nets, but not for the network backingImageCases. Seems there is a lot of MMA glue code to get inference off the ground, which is what makes this a worthy question.

You can access it as follows:

NetModel["Wolfram ImageCases Net V1"]

enter image description here

  • $\begingroup$ I don't believe the ImageCases network is published, but, from a question I put during a livestream, it's a variant of YOLO underneath. I would recommend just using a version of YOLO native to Tensorflow. $\endgroup$
    – Carl Lange
    Jun 11 '19 at 15:55
  • $\begingroup$ @CarlLange it's undocumented, but there $\endgroup$
    – M.R.
    Jun 11 '19 at 19:23
  • 1
    $\begingroup$ Thanks for adding it to your question. And indeed, it doesn't seem to be exactly like YOLO. $\endgroup$
    – Carl Lange
    Jun 11 '19 at 19:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.