Title:
ABNORMAL ENTITY DETECTION SYSTEM AND METHOD
Document Type and Number:
WIPO Patent Application WO/2020/017814
Kind Code:
A1
Abstract:
An abnormal entity detection system according to an embodiment of the present invention comprises: an abnormal entity detection apparatus for detecting an abnormal entity by applying an image including information about an entity to a deep-learning-based first abnormal entity detection model, and, if, as a result of the the abnormal entity detection, it is determined that an error event has occurred, transmitting, to a user terminal, a first error event occurrence image in which the error event occurred; and a training server for receiving error review information corresponding to the first error event occurrence image, and training a deep-learning-based second abnormal entity detection model on the basis of the first error event occurrence image and the error review information, wherein the abnormal entity detection apparatus updates the first abnormal entity detection model on the basis of the training result of the second abnormal entity detection model.
More Like This:
Inventors:
KIM KEUN WOONG (KR)
SHIN JE YONG (KR)
SHIN JE YONG (KR)
Application Number:
PCT/KR2019/008473
Publication Date:
January 23, 2020
Filing Date:
July 10, 2019
Export Citation:
Assignee:
LG INNOTEK CO LTD (KR)
International Classes:
G06T7/00; G06T7/11
Foreign References:
| KR101789690B1 | 2017-10-25 | |||
| KR101830056B1 | 2018-02-19 | |||
| KR20180040287A | 2018-04-20 | |||
| KR20170061016A | 2017-06-02 | |||
| KR20140119228A | 2014-10-10 |
Attorney, Agent or Firm:
DANA PATENT LAW FIRM (KR)
Download PDF: