Granularity-Aware Transfer for Tree Instance Segmentation in Synthetic and Real Forests
Computer Vision and Pattern Recognition
2026-04-16 v1
Abstract
We address the challenge of synthetic-to-real transfer in forestry perception where real data have only coarse Tree labels while synthetic data provide fine-grained trunk/crown annotations. We introduce MGTD, a mixed-granularity dataset with 53k synthetic and 3.6k real images, and a four-stage protocol isolating domain shift and granularity mismatch. Our core contribution is granularity-aware distillation, which transfers structural priors from fine-grained synthetic teachers to a coarse-label student via logit-space merging and mask unification. Experiments show consistent mask AP gains, especially for small/distant trees, establishing a testbed for Sim-Real transfer under label granularity constraints.
Cite
@article{arxiv.2604.13722,
title = {Granularity-Aware Transfer for Tree Instance Segmentation in Synthetic and Real Forests},
author = {Pankaj Deoli and Atef Tej and Anmol Ashri and Anandatirtha JS and Karsten Berns},
journal= {arXiv preprint arXiv:2604.13722},
year = {2026}
}