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

Fine-grained mutual geometric features enhanced human-object interaction detection

delete2026-03-24
delete0
PRE
AI
R
Ri Liu
L
Lin Bai *
S
Shengjie Lin
X
Xiaoyu Jiang
DOI：10.1016/j.patcog.2026.113614delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Propose a novel edge-guided framework to advance HOI detection. • Introduce EGE to encode fine-grained edges, boosting spatial modeling. • Propose EMMA to form robust anchors via median projections of edge points. • Develop ERSA to fuse edge and anchor cues with adaptive spatial attention. • Advance HOI detection by tackling irregular objects and long-range layouts.
Keywords:
edge-guided framework
HOI detection
EGE
EMMA
ERSA

Journal

Pattern Recognition cover
Pattern Recognition
IF:
7.6
Papers:
1.3W
Citations:
4.5W

Organization

No organization information available