Skip to content


ROI pooling for Fast/Faster R-CNN.

This layer performs ROI pooling operation for spatial data. It takes two inputs, input and rois.
The shape of input is (h, w, c).
The shape of rois is (num_roi, 5).
The shape of output is (num_roi, pooled_h, pooled_w, c).


  • pooled_h : The height of pooled output
  • pooled_w : The width of pooled output
  • spatial_scale : Multiplicative spatial scale factor to translate ROI coords from their input scale to the scale used when pooling (Determined by the network structure)