English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Sparse methods and the use of Winograd convolutions are two orthogonalapproaches, each of which significantly accelerates convolution computations inmodern CNNs. Sparse Winograd merges these two and thus has the potential tooffer a combined performance benefit. Nevertheless, training convolution layersso that the resulting Winograd kernels are sparse has not hitherto been verysuccessful. By introducing a Winograd layer in place of a standard convolutionlayer, we can learn and prune Winograd coefficients "natively" and obtainsparsity level beyond 90% with only 0.1% accuracy loss with AlexNet on ImageNetdataset. Furthermore, we present a sparse Winograd convolution algorithm andimplementation that exploits the sparsity, achieving up to 31.7 effectiveTFLOP/s in 32-bit precision on a latest Intel Xeon CPU, which corresponds to a5.4x speedup over a state-of-the-art dense convolution implementation.

Enabling Sparse Winograd Convolution by Native Pruning