Backward pooling computation. More...
#include <include/dlprim/core/pool.hpp>
Inheritance diagram for dlprim::core::Pooling2DBackwardBase:
Public Member Functions | |
| virtual size_t | workspace ()=0 |
| get workspace | |
| virtual void | enqueue (Tensor &X, Tensor &dX, Tensor &dY, float factor, ExecutionContext const &e)=0 |
| X for max pooling to detect the neuron giving max signal. More... | |
Detailed Description
Backward pooling computation.
Member Function Documentation
|
pure virtual |
X for max pooling to detect the neuron giving max signal.
when used with kernel based pooling (not global) X/dX and dY dimensions should match at batch and channels and for H/W the dimention for Y should be Y_dim = op((X_dim + 2 * pad[dim] - kernel[dim]) / stride[dim]) + 1 where op is either ceil or floor
Implemented in dlprim::core::AvgPooling2DBackward.
The documentation for this class was generated from the following file:
- include/dlprim/core/pool.hpp
