0
$\begingroup$

Most pre-trained CNN models accept a $224x224$ input size when they were trained. Can I use $256x256$ to get a higher accuracy?

$\endgroup$
0
$\begingroup$

If you change the image size, you will be able to reuse only part of the original network.

Convolutional and pooling layers can be applied to images of any size, so the initial part of the network, which normally consists of convolutions and pooling, will be reusable as-is.

However, the dense layers after the convolutional part assume certain input dimensions. Therefore, as your larger images lead to larger input to the dense layers, you won't be able to reuse the dense layers.

$\endgroup$
4
  • $\begingroup$ If the answer was helpful, please consider marking it as correct. If the answer is not clear to your, please let me know. $\endgroup$
    – noe
    Jul 21 at 7:02
  • $\begingroup$ can you please explain the answer to this one. Do you agree to it? $\endgroup$
    – djbacs
    Jul 24 at 13:56
  • $\begingroup$ The linked answer is incomplete. If you take a look at the source code link they provide, you will see that the network takes a parameter called “include_top” and only when it is true the network accepts arbitrary shapes. That parameter controls whether you only reuse the convolutional part. This is precisely what I describe in my answer. $\endgroup$
    – noe
    Jul 24 at 17:25
  • $\begingroup$ Please consider marking the answer as correct or, alternative, let me know what is not clear from it. $\endgroup$
    – noe
    Jul 28 at 14:47

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.