Title:
IDENTIFICATION SYSTEM, MODEL RE-LEARNING METHOD AND PROGRAM
Document Type and Number:
WIPO Patent Application WO/2019/215780
Kind Code:
A1
Abstract:
In the present invention, a learning means 701 learns, using teacher data, a model for identifying an object represented by data. A first identification means 702 identifies an object represented by data, using a model learned by the learning means 701. A second identification means 703 identifies, using a different model to that learned by the learning means 701, an object represented by data subjected to identification by the first identification means 702. The learning means 701 re-learns the model using teacher data which includes a label for the data stipulated on the basis of the identification results derived by the second identification means 703, and that data.
More Like This:
| JPH0269828 | SELF LEARNING EXPERT SYSTEM |
| JPH05159066 | IMAGE FORMING DEVICE |
| JPH05297893 | PREPROCESSOR FOR ACOUSTIC SIGNAL |
Inventors:
INOSHITA TETSUO (JP)
Application Number:
PCT/JP2018/017611
Publication Date:
November 14, 2019
Filing Date:
May 07, 2018
Export Citation:
Assignee:
NEC CORP (JP)
International Classes:
G06N99/00; G06V10/25
Foreign References:
| JP2012043225A | 2012-03-01 | |||
| JP2017117139A | 2017-06-29 |
Attorney, Agent or Firm:
IWAKABE Fuyuki et al. (JP)
Download PDF: