Title:
ACCELERATION METHOD AND ACCELERATOR USED FOR CONVOLUTIONAL NEURAL NETWORK
Document Type and Number:
WIPO Patent Application WO/2019/196223
Kind Code:
A1
Abstract:
The present invention provides an acceleration method and accelerator used for a convolutional neural network, the method comprising: S1, for any layer of the convolutional neural network, calculating the density of various feature maps output by this layer; S2, comparing the density of the various feature maps output by this layer to a plurality of preset thresholds, and performing, according to the comparison results, sparse coding to the various feature maps, different comparison results corresponding to different sparse coding schemes; and S3, performing, on the basis of the convolution layer of a layer following this layer, convolution to the various sparse coded feature maps and various pre-sparse coded convolution kernels in the convolutional neural network. The present invention reduces the calculated amount of the convolution operation in the convolutional neural network and improves operation speed.
Inventors:
LIU YONGPAN (CN)
YUAN ZHE (CN)
YUE JINSHAN (CN)
YANG HUAZHONG (CN)
LI XUEQING (CN)
WANG ZHIBO (CN)
YUAN ZHE (CN)
YUE JINSHAN (CN)
YANG HUAZHONG (CN)
LI XUEQING (CN)
WANG ZHIBO (CN)
Application Number:
PCT/CN2018/095365
Publication Date:
October 17, 2019
Filing Date:
July 12, 2018
Export Citation:
Assignee:
UNIV TSINGHUA (CN)
International Classes:
G06N3/04
Foreign References:
CN107239825A | 2017-10-10 | |||
CN107239824A | 2017-10-10 | |||
CN105184362A | 2015-12-23 | |||
CN107609641A | 2018-01-19 | |||
US20170103309A1 | 2017-04-13 |
Attorney, Agent or Firm:
CN-KNOWHOW INTELLECTUAL PROPERTY AGENT LIMITED (CN)
Download PDF: