Ideally I wouldn't need to do training for Watson to recognize my logo, but if it is needed, I wouldn't mind uploading a small number of images of my logo (<10 images) to train the system. Watson should be able to identify multiple instances of the logo within an image and provide me with bounding boxes/locations of where they appear.
Why is it useful?
|Who would benefit from this IDEA?||As a user, I want to be able to identify my company's logo within images.|
How should it work?