Title:
SEMICONDUCTOR DEVICE HAVING NEURAL NETWORK
Document Type and Number:
WIPO Patent Application WO/2018/234919
Kind Code:
A1
Abstract:
Provided is a semiconductor device capable of efficiently performing image recognition using a neural network. The semiconductor device comprises: a shift register group; a D/A converter; and a multiply-accumulate circuit. The multiply-accumulate circuit comprises an analog memory and stores filter parameters. The shift register group captures image data and outputs some image data to the D/A converter while shifting the image data. The D/A converter converts some of the input image data into analog data and outputs the analog data to the multiply-accumulate circuit.
More Like This:
WO/2019/165432 | EXTENDED REALITY MEDICAL REPORT GENERATING SYSTEM |
JPH0467256 | LEARNING SYSTEM FOR NEURAL NETWORK |
JPH0782483 | [Title of Invention] Neural network learning device |
Inventors:
KUROKAWA YOSHIYUKI (JP)
HARADA SHINTARO (JP)
HARADA SHINTARO (JP)
Application Number:
PCT/IB2018/054188
Publication Date:
December 27, 2018
Filing Date:
June 11, 2018
Export Citation:
Assignee:
SEMICONDUCTOR ENERGY LAB (JP)
International Classes:
G06G7/60; G06N3/063; H01L29/786
Foreign References:
JP2010134697A | 2010-06-17 | |||
JP2017054502A | 2017-03-16 | |||
JP2008113295A | 2008-05-15 |
Download PDF: