Contributions
- eliminate all the redundant computation in convolution and pooling on images by introducing novel d-regularly sparse kernels.
- It generates exactly the same results as those by patch-by-patch scanning.
- The proposed d-regularly sparse kernels not only ensure exactly the same results as patch-bypatch scanning in both forward and backward propagation, but also allow to access memory in a continuous manner, which is the key to fully utilize the computational capability of GPUs, regardless of the strides of convolution and pooling in the original CNN
- By applying a mask to the error map at the last layer of a CNN, one can choose an arbitrary subset of patches of interest from a training image to update CNN parameters via backward propagation and with constant computation complexity.
Methods
- The Key of Our Approach:modify both the convolution and pooling kernels of the original CNN by inserting a specific number of all-zero columns and rows to compensate for the down-sampling by the convolution and pooling layers.
Architecture
Others
- Directly applying it to pixelwise classification in a patch-by-patch scanning manner is extremely inefficient, because surrounding patches of pixels have large overlaps, which lead to a lot of redundant computation