Real-world uses

Created for this library

1. 1.

   A computer vision team uses spatial pooling in its CNN to downsample feature maps while preserving dominant activations.

2. 2.

   A medical imaging team uses spatial pooling at the end of its CNN to produce a fixed-size representation for the classifier head.

3. 3.

   A research team uses spatial pyramid pooling so its detector handles inputs of varying sizes.