Title:
出力方法、出力プログラム、および出力装置
Document Type and Number:
Japanese Patent JP7207568
Kind Code:
B2
Abstract:
A computer-implemented outputting method including: generating a correction vector that corrects a vector based on information of a first modal on the basis of correlation between the vector based on the information of the first modal and a vector based on information of a second modal; combining the generated correction vector with the vector based on the information of the first modal; compressing the combined vector based on the information of the first modal according to a predetermined rule; performing normalization processing for the compressed vector based on the information of the first modal; and outputting a vector obtained by the normalization processing.
Inventors:
Moe Yamada
Application Number:
JP2021555729A
Publication Date:
January 18, 2023
Filing Date:
November 14, 2019
Export Citation:
Assignee:
富士通株式会社
International Classes:
G06N20/00; G06T7/00
Other References:
LU, Jiasen, et al.,ViLBERT: Pretraining task-agnostic visiolinguistic representations for vision-and-language tasks,arXiv.org [online],2019年08月06日,pp.1-11,[検索日 2019.12.13], インターネット:
NGUYEN, Duy-Kien, et al.,Improved fusion of visual and language representations by dense symmetric co-attention for visual qu,[online],2018年,pp.6087-6096,http://openaccess.thecvf.com/content_cvpr_2018/html/Nguyen_Improved_Fusion_of_CVPR_2018_paper.html,[検索日 2019.12.13], インターネット:
NGUYEN, Duy-Kien, et al.,Improved fusion of visual and language representations by dense symmetric co-attention for visual qu,[online],2018年,pp.6087-6096,http://openaccess.thecvf.com/content_cvpr_2018/html/Nguyen_Improved_Fusion_of_CVPR_2018_paper.html,[検索日 2019.12.13], インターネット:
Attorney, Agent or Firm:
Akinori Sakai