Given that I have a set of images that contain multiple objects for which labels exist and the object the image label refers to is always in the center. The objects vary in size. I want to train a convolutional neural network. Would it be a good idea to blur the image radially increasing? If not, what preprocessing or data augmentation would be appropriate in such a situation?

I haven't done an experiment yet. I'm more interested in a theoretical discussion of the effects.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.