Toward a foundation model for forest point clouds

Yuanwen Yue, Stefano Puliti, Damien Robert, Atakan Topaloğlu, Binbin Xiang, Maciej Wielgosz, Jan Dirk Wegner, Rasmus Astrup, Christian Rupprecht, Konrad Schindler · arXiv 2026 · 2026

arXiv Code Project page

Status

To read

Topics

backbones, forest-segmentation

Feeds

First paper · height sampling

What they did

Takes a step toward a foundation model for forest point clouds: with LitePT as the backbone, it asks whether one pretrained model can learn representations that transfer across forest inventory tasks, sensors and forest types.

My note

No note yet.

Code

Repository

prs-eth/ForPT

Framework

PyTorch

Licence

MIT

My status

Not tried

BibTeX

@article{yue2026-forpt,
  title = {{Toward a foundation model for forest point clouds}},
  author = {Yuanwen Yue and Stefano Puliti and Damien Robert and Atakan Topaloğlu and Binbin Xiang and Maciej Wielgosz and Jan Dirk Wegner and Rasmus Astrup and Christian Rupprecht and Konrad Schindler},
  journal = {arXiv 2026},
  year = {2026},
  eprint = {2609.24787},
  archivePrefix = {arXiv}
}