Abstract: Winograd algorithm can effectively reduce the computational complexity of convolution operation. Effectively using the parallelism of Winograd convolution algorithm can effectively improve ...

Winograd's fast convolution algorithms transform input and filters into another space where convolution becomes element-wise multiplication. The fourier transform also turns convolutions into ...

ABSTRACT: The first error theory and bounds for Fast Matrix Multiplication based on the Strassen-Winograd algorithms (FastMMW) were formulated in the 70s. The theory ...

Recently, small convolutional filter sizes have become an important component in convolutional neural networks such as Google’s AlphaGo network or Microsoft’s deep residual networks. While most ...