0 Replies
      Latest reply on Feb 28, 2020 7:47 AM by ORStudiosUK
      ORStudiosUK Level 1 Level 1 (0 points)

        I am new to ML and I am a little confused about how to go about training the model with images that contain multiple objects. Lets say for example I have an image of a lake and it contains water, mountains, sky. I would like the model to recognise each of these objects in the photo so do I simply copy the same image into classes for water, mountains etc?

         

        I have tried this with around 30-50 images and the accuracy is down around the 20-30% range.

         

        I am trying to expand a photo editing app I already have to include detection of nightime, daytime etc.

         

        Any help/guidance on this would be great thanks.