SPP layer could be added in CNN model between convolutional layer and fully-connected lay, so that you can input multi-size images into your CNN model.
The function spatial_pyramid_pool()
in file spp_layer.py
is independent. It could be added in your own models.
See this:Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
yinizhizhu/sppnet-pytorch
A simple Spatial Pyramid Pooling layer which could be added in CNN
PythonApache-2.0