Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
PRE-PROCESSING DEVICE, PRE-PROCESSING METHOD, AND PRE-PROCESSING PROGRAM
Document Type and Number:
WIPO Patent Application WO/2020/116129
Kind Code:
A1
Abstract:
In a learning device (10), a pre-processing unit (130) has: a pre-learning data collection unit (131) that collects, as pre-learning data, continuous input data measured under an environment simulating an estimated environment and output data corresponding to the continuous input data; and a conversion unit (132) that converts the continuous input data into continuous input data having multiple sizes including a size larger than that of the input data, converts the output data corresponding to the continuous input data into output data respectively corresponding to the continuous input data having multiple sizes, and outputs the output data as learning data.

Inventors:
YAMASHITA JUMPEI (JP)
KOYA HIDETAKA (JP)
NAKAJIMA HAJIME (JP)
Application Number:
PCT/JP2019/044884
Publication Date:
June 11, 2020
Filing Date:
November 15, 2019
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
NIPPON TELEGRAPH & TELEPHONE (JP)
International Classes:
G06N20/00; G06N3/08
Foreign References:
JP2017091278A2017-05-25
Other References:
SASAKI, KAZUMA ET AL.: "Handling the simulation-to-reality gap for learning systems of real robots: A survey", IEICE TECHNICAL REPORT, vol. 118, no. 260, 12 October 2018 (2018-10-12), pages 5 - 8
MALHOTRA, PANKAJ: "TimeNet: Pre-trained deep recurrent neural network for time series classification", ARXIV.ORG, 2017, XP080772899
Attorney, Agent or Firm:
SAKAI INTERNATIONAL PATENT OFFICE (JP)
Download PDF: