I have a couple of questions. Im wondering if anyone has used Clarifiai to Classifiy human actions/poses?
I have been using Keras to build a custom classifier to i am pretty excited looking at Clarifai; i'm more interested in actions of hands around the face so it looks like the face detection or age gender ethnicity models could be trained using some of my own images?
Does custom training take place through fine-tuning? It feels like 10 images for a concept is a low number (though if it works from 10 images thats incredible!).
Also i'm wondering if it's possible to export trained models and weights - i'm guessing not but its worth asking!
Edit: so i see that the models i am interested in (i.e. face detection) are not available yet when making an application? Is there any indication on when they will be OR is it possible for me to get early access in exchange for some feedback??