How to handle dynamic input size in neural networks

What ways are there to handle dynamic input size for neural networks?

neural-networkdynamic-size
2 votesLP185.00
1 Answers
NN132.00
1

There are 3 solutions for this

  1. Resize all images to one common size
  2. Use 2D padding for images smaller than input size (resize all other images, which are bigger)
  3. Use RoI align before you get the top of the network
Reply
Couldn't find what you were looking for?and we will find an expert to answer.
How helpful was this page?