0
$\begingroup$

i'm trying to classify image pattern with CNN; I started to optimize a neural network with image represented in cartesian coordinate. If I use image represented in polar coordinate should i totaly change my neural network?

Since my image will have the same size, but characteristic will change (for exemple density in an area of the image will change for a given pattern in the new representation)

Improve this question
$\endgroup$
1
  • $\begingroup$ What do you means by in polar coordinates? Can you show example $\endgroup$
    – Jon Nordby
    Commented May 5, 2019 at 1:04

1 Answer 1

Reset to default
0
$\begingroup$

Whenever the data distribution changes, you need to retrain your neural network. However it could very well be that you can keep the same model architecture. You can also try to keep the existing weights from the old data (transfer learning), it might perform better than training from scratch. Or it can be worse.

Improve this answer
$\endgroup$

Your Answer

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

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