Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
NEURAL NETWORK GENERATION DEVICE, NEURAL NETWORK COMPUTING DEVICE, EDGE DEVICE, NEURAL NETWORK CONTROL METHOD, AND SOFTWARE GENERATION PROGRAM
Document Type and Number:
WIPO Patent Application WO/2022/163861
Kind Code:
A1
Abstract:
This neural network generation device generates a neural network execution model that computes a neural network. The neural network execution model converts input data containing an element of 8 bits or more to a converted value with less bits than the element, on the basis of comparison with a plurality of threshold values.

Inventors:
TOKUNAGA HIROYUKI (JP)
Application Number:
PCT/JP2022/003745
Publication Date:
August 04, 2022
Filing Date:
February 01, 2022
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
LEAPMIND INC (JP)
International Classes:
G06N3/02
Other References:
IMAI, TAKUJI: "Power efficiency improvement of DNN inference by large-scale quantization to 1 to 2 bits Leap Mind develops CNN accelerator circuit", NIKKEI ROBOTICS, no. 60, 1 July 2020 (2020-07-01), pages 12 - 17, XP009538495, ISSN: 2189-5783
IMAI, TAKUJI: "Sony learns the optimal number of bits in deep neural net's new quantization technology", NIKKEI ROBOTICS, no. 61, 1 August 2020 (2020-08-01), pages 16 - 22, XP009538498, ISSN: 2189-5783
SHIMODA MASAYUKI, SATO SHIMPEI, NAKAHARA HIROKI: "All binarized convolutional neural network", 2017 INTERNATIONAL CONFERENCE ON FIELD PROGRAMMABLE TECHNOLOGY (ICFPT), IEICE, JP, 1 December 2017 (2017-12-01) - 14 September 2017 (2017-09-14), JP, pages 51 - 57, XP055954726, DOI: 10.1109/FPT.2017.8280163
Attorney, Agent or Firm:
NISHIZAWA Kazuyoshi et al. (JP)
Download PDF: