.png){kind=icon}
.png){kind=icon}
Return
Label-efficient structure-aware learning framework for wheat tiller density estimation from UAV remote sensing
.png){kind=icon}
PRE
AI
M
Y
T
Z
L
M
M
DOI：10.1016/j.compag.2026.112284.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Label-efficient framework for UAV-based wheat tiller density estimation. • Contrastive learning separates crops without pixel-level annotations. • Self-distillation learns structure-sensitive canopy features from 2D imagery. • Synergistic fusion models structure-physiology complementarity. • Robust cross-field small-sample adaptation with limited target-field labels.
Keywords:
Tiller density estimation
UAV remote sensing
Structure-aware
Self-supervised learning
Label-efficient learning
Cross-field adaptation
Journal
IF:
8.9
Papers:
9.9K
Citations:
4.8W
Organization
Cited Papers