SEMIR uses topology-preserving graph minors to collapse pixel lattices into small supernode graphs for GNN classification, enabling full-resolution thin-structure segmentation without patching and matching baselines on three datasets with 4.6x less fragmentation.
arXiv preprint arXiv:2304.05821 (2023)
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
-
SEMIR: Topology-Preserving Graph Minors for Thin-Structure Segmentation
SEMIR uses topology-preserving graph minors to collapse pixel lattices into small supernode graphs for GNN classification, enabling full-resolution thin-structure segmentation without patching and matching baselines on three datasets with 4.6x less fragmentation.