Published August 5, 2025 | Version v1
Dataset Open

FOR-instanceV2 dataset and pre-trained model in paper entitled "ForestFormer3D: A Unified Framework for End-to-End Segmentation of Forest LiDAR 3D Point Clouds"

Creators

  • 1. ROR icon Norwegian Institute of Bioeconomy Research
  • 2. Norwegian Institute for Bioeconomy Research
  • 3. ROR icon Silva Tarouca Research Institute for Landscape and Ornamental Gardening

Files

clean_forestformer.zip

Files (3.0 GB)

Name Size Download all
md5:553d67379331966509076f3fbb409e57
197.8 MB Preview Download
md5:1c00a0f0b89f03b74064432162619136
377.1 MB Preview Download
md5:5a63cc1cbe88edd9ebec28ad7e46f79b
2.4 GB Preview Download

Additional details

Identifiers

arXiv
arXiv:2506.16991

Related works

Is described by
Conference proceeding: arXiv:2506.16991 (arXiv)

Funding

The Research Council of Norway
Bringing Industry 4.0 to the Norwegian forest sector 309671
Czech Academy of Sciences
INTER‑EXCELLENCE INTER‑COST LUC23023 “A step towards advanced forest inventory through laser scanning and 3D modelling. Can field surveys be minimized?” LUC 23023

Dates

Available
2025-08-05

References

  • ICCV 2025 paper: ForestFormer3D: A Unified Framework for End-to-End Segmentation of Forest LiDAR 3D Point Clouds