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

Segmentation-guided association refinement in multiple object tracking

delete2026-09-18
delete0
delete
OA
AI
M
Manuel Bendaña *
V
V.M. Brea
M
Manuel Mucientes
DOI：10.1016/j.neucom.2026.135176delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Tracking by Detection (TbD) fails under occlusion; segmentation cues enhance recovery. • SGAR-MOT boosts TbD approaches with mask-based reasoning to reinstate lost tracks. • ORP recovers unmatched tracks using SAM 2 and a learned identity module. • MIR verifies identity consistency from multi-frame segmentation evidence. • SGAR-MOT reduces errors vs. baseline and achieves, on average, the best SoTA results.
Keywords:
Multiple object tracking
Tracking-by-detection
Segmentation-guided object recovery

Journal

Neurocomputing cover
Neurocomputing
IF:
6.5
Papers:
2.5W
Citations:
6.5W

Organization

No organization information available
Cited Papers

Cited Papers

No cited papers available