The paper introduces SegmentAnyTreeV2, a new framework using Point Transformer v3 for tree instance segmentation in point clouds, along with the FOR-instance v3 benchmark of 427 scenes and 26,496 trees, achieving high performance metrics and good generalization.
RayExtract: A fast, scalable method for tree volume reconstruction from terrestrial laser scanning
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
-
SegmentAnyTreeV2: Scaling Transformer-Based Tree Instance Segmentation Across Sensors, Platforms, and Forests
The paper introduces SegmentAnyTreeV2, a new framework using Point Transformer v3 for tree instance segmentation in point clouds, along with the FOR-instance v3 benchmark of 427 scenes and 26,496 trees, achieving high performance metrics and good generalization.