MATLAB Answers

0

Transfer learning without imresize or imageDatastore

Asked by Michael Benton on 12 Aug 2018 at 20:24
Latest activity Edited by Michael Benton on 12 Aug 2018 at 21:09

For many samples of small images, it would be nice to load the data into RAM and perform transfer learning without using images on slower memory (imageDatastore). Unfortunately, if I had 50x50x1 images and had to resize to go with alexnet, I will be forced to use (50x50x1/227x227x3) ~ 1.6% of the number of samples in order to keep everything in RAM.

Does anyone know a fix? A custom layer that resizes would work, that's a lot of work though.

using 2017b

  0 Comments

Sign in to comment.

0 Answers