keep the resize function the same in test time the same with training time

[hcx1231]

The result may be affected when the implementation of resize function in inference time and training time is inconsistent,so I changed the code in inference time to keep them same.

[facebook-github-bot]

Thank you for your pull request and welcome to our community. We require contributors to sign our Contributor License Agreement, and we don't seem to have you on file. In order for us to review and merge your code, please sign up at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need the corporate CLA signed.

If you have received this in error or have any questions, please contact us at cla@fb.com. Thanks!

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Facebook open source project. Thanks!

[fmassa]

The only difference I see is that the images might be larger than 1333 pixels, but is this a problem?

[hcx1231]

The only difference I see is that the images might be larger than 1333 pixels, but is this a problem?

Yes,when the images are very large,for example in my own dataset,it may cause large error as it haven't been trained in such size,which took me several weeks to find out this problem.

[fmassa]

Ok, fair enough.