Adaptive Tiling: Applying Fixed-size Systolic Arrays to Sparse Convolutional Neural Networks

H. T. Kung, Bradley McDanel, Sai Qian Zhang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We introduce adaptive tiling, a method of partitioning layers in a sparse convolutional neural network (CNN) into blocks of filters and channels, called tiles, each implementable with a fixed-size systolic array. By allowing a tile to adapt its size so that it can cover a large sparse area, we minimize the total number of tiles, or equivalently, the number of systolic array calls required to perform CNN inference. The proposed scheme resolves a challenge of applying systolic array architectures, traditionally designed for dense matrices, to sparse CNNs. To validate the approach, we construct a highly sparse Lasso-Mobile network by pruning MobileNet trained with an ℓ1 regularization penalty, and demonstrate that adaptive tiling can lead to a 2- 3x reduction in systolic array calls, on Lasso-Mobile, for several benchmark datasets.

Original languageEnglish (US)
Title of host publication2018 24th International Conference on Pattern Recognition, ICPR 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1006-1011
Number of pages6
ISBN (Electronic)9781538637883
DOIs
StatePublished - Nov 26 2018
Event24th International Conference on Pattern Recognition, ICPR 2018 - Beijing, China
Duration: Aug 20 2018Aug 24 2018

Publication series

NameProceedings - International Conference on Pattern Recognition
Volume2018-August
ISSN (Print)1051-4651

Other

Other24th International Conference on Pattern Recognition, ICPR 2018
Country/TerritoryChina
CityBeijing
Period8/20/188/24/18

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Adaptive Tiling: Applying Fixed-size Systolic Arrays to Sparse Convolutional Neural Networks'. Together they form a unique fingerprint.

Cite this