Title:
METHOD AND SYSTEM FOR ESTIMATING MOTION OF WORK VEHICLE, METHOD FOR PRODUCING LEARNED CLASSIFICATION MODEL, LEARNING DATA, AND METHOD FOR PRODUCING LEARNING DATA
Document Type and Number:
WIPO Patent Application WO/2019/093152
Kind Code:
A1
Abstract:
This method is executed by a computer and includes the following processing. First processing acquires motion data. The motion data is generated from a plurality of images showing a work vehicle in motion in time series and shows a motion change of the work vehicle. Second processing determines the classification of the motion of the work vehicle from the motion data by executing image classification using a learned classification model.
Inventors:
AIZAWA MASANORI (JP)
Application Number:
PCT/JP2018/039830
Publication Date:
May 16, 2019
Filing Date:
October 26, 2018
Export Citation:
Assignee:
KOMATSU MFG CO LTD (JP)
International Classes:
G06T7/00
Foreign References:
| JP2002298143A | 2002-10-11 | |||
| KR101441107B1 | 2014-09-23 | |||
| JPH10290450A | 1998-10-27 | |||
| JPH1060948A | 1998-03-03 |
Other References:
HASHIMURA, KEISUKE ET AL.: "Sign language recognition by convolutional neural network with concatenated sequence image of depth image", IEICE TECHNICAL REPORT., vol. 116, no. 248, 9 October 2016 (2016-10-09), pages 17 - 22, XP055607222
SHIMIZU, RYO: "Distillation'' for Dramatically Raising the Efficiency of Deep Neural Networks", WIRELESSWIRE NEWS, 30 September 2016 (2016-09-30), Retrieved from the Internet [retrieved on 20181217]
SHIMIZU, RYO: "Distillation'' for Dramatically Raising the Efficiency of Deep Neural Networks", WIRELESSWIRE NEWS, 30 September 2016 (2016-09-30), Retrieved from the Internet
Attorney, Agent or Firm:
SHINJYU GLOBAL IP (JP)
Download PDF:
Previous Patent: SOLID-STATE IMAGE PICKUP DEVICE AND ELECTRONIC APPARATUS
Next Patent: CELL BALANCE CONTROL DEVICE, AND CELL BALANCE CONTROL METHOD
Next Patent: CELL BALANCE CONTROL DEVICE, AND CELL BALANCE CONTROL METHOD