logo
CanyamAI summaries for academic research
language
MobileMobile
  • Phone FeatureWeChat Mini Program
EnglishEnglish
noticeNotice
BlogBlog
delete
arrow
Return

GNN-based primitive recombination for compositional zero-shot learning

delete2025-10-08
delete0
PRE
AI
F
Fuqin Deng
C
C. J. Tang
L
Lanhui Fu
W
Wei Jin
J
Jiaming Zhong
H
Hongming Wang
N
N Li *
DOI：10.1016/j.imavis.2025.105762delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• We propose a GNN-based Primitive Recombination (GPR) framework. • We design a graph model to enhance interaction among visual primitives. • Primitive-level CLIP loss is introduced to optimize visual feature.

Journal

Image and Vision Computing cover
Image and Vision Computing
IF:
4.2
Papers:
4.0K
Citations:
6.7K

Organization

M
Macau University of Science and Technology
Scholars:
2.4K
Papers: 1.4K
Citations: 9.7K
W
Wuyi University
Scholars:
4.3K
Papers: 2.3K
Citations: 2.9K
G
Guangdong University of Science and Technology
Scholars:
166
Papers: 100
Citations: 0
researcher View more organizations