.png){kind=icon}
.png){kind=icon}
Return
Two-step distance-based skydiving formation recognition
.png){kind=icon}
PRE
AI
A
O
DOI：10.1016/j.imavis.2026.106085.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Machine learning can assist approximate scoring in formation skydiving. • A two-step pose-based skydiving formation classifier is proposed. • Keypoint-only, distance-based, and wrist distance-based variants evaluated. • Keypoint-only variant fails to classify reliably (weighted F1 = 0.64). • Wrist distance + FCNN achieves F1 = 0.95, matching a single-step baseline.
Journal
IF:
4.2
Papers:
4.0K
Citations:
6.7K
Organization
No organization information available
Cited Papers