Generally speaking, what's the best way to approach an image classification task with unevenly sized images (e.g. example 1 has size 300x240, example 2 has size 240x224 etc.)? Ideally I would like the dimensions to be consistent across all images before commencing CNN training.

I'm quite new to deep learning and would really appreciate any help.



The size of the input image is usually decided by computational resources you have. If input image has larger dimension you need to increase filter size, then model will have more parameters, hence you need more computational resources and longer training time.

Usually, I resize the images to min(heights of all images), min(widths of all images).

If there are larger samples of particular dimension in the dataset and you have enough computational resource, it might also makes sense to resize to that dimension.

Try with different settings, pick the one that best works for you.


Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.