Pascoassis Souza Santos Meira, Gilson Antonio Giraldi, Rafael Duarte Coelho dos Santos · 2026 · arXiv (Cornell University)
DOI: 10.48550/arxiv.2610.30007 · cited by 0 works (OpenAlex) · open-access PDF
La segmentation sémantique des images de détection à distance place souvent les limites de classe à quelques pixels de là où elles se trouvent réellement, et cette erreur de position est concentrée dans la bande mince mesurée par l'IoU de frontière (BIoU). Nous traitons le raffinement de frontière comme une application itérée d'une petite règle locale apprise: le Boundary Refinement Neural Cellular Automaton (BRNCA), qui commence à partir des probabilités de classe d'un segmentateur congelé, pré-entraîné et les corrige cellule par cellule. Nous généralisons le rNCA binaire au cas multiclasse en gardant les canaux visibles sur le simplex de probabilité à travers une mise à jour Aitchison (log-space) et en ajoutant des c
Machine translation (Meta M2M100); the English original follows.
Semantic segmentation of remote sensing imagery often places class boundaries a few pixels away from where they truly lie, and this position error is concentrated in the thin band measured by Boundary IoU (BIoU). We treat boundary refinement as the iterated application of a small learned local rule: the Boundary Refinement Neural Cellular Automaton (BRNCA), that starts from the class probabilities of a frozen, pre-trained segmenter and corrects them cell by cell. We generalise the binary rNCA to the multiclass case by keeping the visible channels on the probability simplex through an Aitchison (log-space) update and by adding free hidden channels. The best configuration improves the mean BIoU over the base segmenter by 2.36 points (from 46.63 to 48.99). The gain is concentrated on the irregular woodland class (+8.5 points BIoU); this suggests that position error is correlated with the fractal dimension of the class boundaries.
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
Topics: Boundary (topology), Computer science, Segmentation, Artificial intelligence, Remote sensing, Pixel
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.30007, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange