Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
CONDENSE-EXPANSION-DEPTH-WISE CONVOLUTIONAL NEURAL NETWORK FOR FACE RECOGNITION
Document Type and Number:
WIPO Patent Application WO/2020/051816
Kind Code:
A1
Abstract:
Techniques related to implementing convolutional neural networks for face or other object recognition are discussed. Such techniques may include applying, in turn, a depth-wise separable convolution, a condense point-wise convolution, and an expansion point-wise convolution to input feature maps to generate output feature maps such that the output from the expansion point-wise convolution has more channels than the output from the condense point-wise convolution.

Inventors:
CHEN YURONG (CN)
LI JIANGUO (CN)
Application Number:
PCT/CN2018/105380
Publication Date:
March 19, 2020
Filing Date:
September 13, 2018
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
INTEL CORP (US)
CHEN YURONG (CN)
LI JIANGUO (CN)
International Classes:
G06T7/00; G06N3/04
Domestic Patent References:
WO2018003212A12018-01-04
Foreign References:
US9436895B12016-09-06
CN107316015A2017-11-03
CN108171112A2018-06-15
Other References:
HOWARD, ANDREW G. ET AL.: "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications", ARXIV:1704.04861V1, 17 April 2017 (2017-04-17), XP080763381, Retrieved from the Internet
XIE, SAINING ET AL.: "Aggregated Residual Transformations for Deep Neural Networks", ARXIV:1611.05431V2, 16 November 2016 (2016-11-16), pages 3, XP080732315, Retrieved from the Internet
OUPPAPHAN PICHAYOOT: "2017 21 ST INTERNATIONAL COMPUTER SCIENCE AND ENGINEERING CONFERENCE (ICSEC", IEEE, article "Corn Disease Identification from Leaf Images Using Convolutional Neural Networks"
ZHAO RUIZHE ET AL.: "2018 28TH INTERNATIONAL CONFERENCE ON FIELD PROGRAMMABLE LOGIC AND APPLICATIONS (FPL", IEEE, article "Towards Efficient Convolutional Neural Network for Domain-Specific Applications on FPGA"
See also references of EP 3850580A4
Attorney, Agent or Firm:
NTD PATENT AND TRADEMARK AGENCY LTD. (CN)
Download PDF: