I bought an issue with the output of my MLModel. For instance I wish to create a mannequin with solely two categorys, canines and cats and nothing else. I’m handing over an information folder with two sub folders, canine and cat.The canine folder such because the cat folder will get the identical quantity of images. Similar for the check folder. After I embed the mannequin into my Xcode and I’m making an attempt to take an image of a canine or a cat every little thing works tremendous, BUT when I attempt to make an image of for instance a keyboard I get “canine” as output with a confidence of over 99%. I simply wish to output unknown object as a substitute of one in every of these two categorys.
I additionally tried to recreate the mannequin with extra loops and additionall augmentations however the consequence continues to be the identical. I thinked of mayybe a default folder within the information folder and a check folder for different circumstances than canines and cats,however I dont assume that that is the precise means for fixing the issue.
