Granularity-Aware Transfer for Tree Instance Segmentation in Synthetic and Real Forests
For forestry perception researchers, this provides a testbed and method for Sim-Real transfer under label granularity constraints.
The paper tackles synthetic-to-real transfer for tree instance segmentation when real data have only coarse labels. Their granularity-aware distillation method achieves consistent mask AP gains, especially for small/distant trees.
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.