For the task of detecting an object of any color, if we train our CNN only with images that only contain that object in one or two colors, will the accuracy of our model's predictions be affected for images containing that object with different colors than the ones used on training?

If so, what is the best way to attain color invariance? (Examples and references to credible sources would be much appreciated.)

  • $\begingroup$ one way would be to train with color variations of images (eg create alternative train images from same image where the object is colored differently), another way would be to normalise the images into some definite colors and train and test with them $\endgroup$ – Nikos M. May 9 at 8:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.