activity
20152023
most citedLow-memory GEMM-based convolution algorithms for deep neural networks

44 citations · 167 across the 26 of their papers we have counts for

collaborators
Showing 2018 · math.NAShow all

1 paper · 2 filters

math.NA2018

Error Analysis and Improving the Accuracy of Winograd Convolution for Deep Neural Networks

Barbara Barabasz, Andrew Anderson, Kirk M. Soodhalter +1

Popular deep neural networks (DNNs) spend the majority of their execution time computing convolutions. The Winograd family of algorithms can greatly reduce the number of arithmetic…