Title:
MODEL GENERATION METHOD, MODEL GENERATION DEVICE, INFERENCE PROGRAM, AND INFERENCE DEVICE
Document Type and Number:
WIPO Patent Application WO/2023/140044
Kind Code:
A1
Abstract:
This model generation device executes machine learning of an inference model comprising a preprocessing module and a graph inference module. The preprocessing module comprises a feature extractor and a selection module. The feature extractor calculates the feature amount of each element belonging to each of a plurality of sets included in an input graph. The selection module selects one or more branches extending from each element on the basis of the calculated feature amount of each element, and generates, for each set, graph information indicating the calculated feature amount of each element and the branch selection result. The graph inference module is differentiable, and infers a task solution with respect to an input graph from the graph information generated for each set.
More Like This:
Inventors:
YANAGI RINTARO (JP)
HASHIMOTO ATSUSHI (JP)
CHIBA NAOYA (JP)
USHIKU YOSHITAKA (JP)
HASHIMOTO ATSUSHI (JP)
CHIBA NAOYA (JP)
USHIKU YOSHITAKA (JP)
Application Number:
PCT/JP2022/047490
Publication Date:
July 27, 2023
Filing Date:
December 22, 2022
Export Citation:
Assignee:
OMRON TATEISI ELECTRONICS CO (JP)
International Classes:
G06N3/09; G06N3/04
Foreign References:
| US20210232918A1 | 2021-07-29 |
Other References:
KEXUE FU; SHAOLEI LIU; XIAOYUAN LUO; MANNING WANG: "Robust Point Cloud Registration Framework Based on Deep Graph Matching", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, 7 March 2021 (2021-03-07), 201 Olin Library Cornell University Ithaca, NY 14853 , XP081907454
Attorney, Agent or Firm:
IP FIRM SHUWA (JP)
Download PDF: