Hi
I am currently replicating your pipeline as part of my analysis, I noticed that for training, data is standardised with the mean/std of pictures within dataset, whereas for testing, the pixel values are standardised per image. Can you please let me know why the standardisation process is different between training and testing and if there is a special reason?
Thanks
Yuju
Hi
I am currently replicating your pipeline as part of my analysis, I noticed that for training, data is standardised with the mean/std of pictures within dataset, whereas for testing, the pixel values are standardised per image. Can you please let me know why the standardisation process is different between training and testing and if there is a special reason?
Thanks
Yuju