Extends Mask R-CNN with attribute regression branches and custom losses to predict roof masks plus height, slope, and azimuth from aerial images, reporting MAEs of ~1m, 4°, 7° on Dutch data with AP50 of 0.566.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Joint Instance Segmentation and Geometric Attribute Regression for Roof Structures in Aerial Imagery
Extends Mask R-CNN with attribute regression branches and custom losses to predict roof masks plus height, slope, and azimuth from aerial images, reporting MAEs of ~1m, 4°, 7° on Dutch data with AP50 of 0.566.