Training-Free Fine-Grained Semantic Segmentations in Low Data Regimes: A FungiTastic Baseline

Sebastian Cavada, Francesco Pelosin, Lapo Faggi · 2026

Read Training-Free Fine-Grained Semantic Segmentations in Low Data Regimes: A FungiTastic Baseline by Sebastian Cavada, Francesco Pelosin, Lapo Faggi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-22492.txt.

Description

Cavada, Pelosin and Faggi propose a training-free two-stage framework for mushroom segmentation in low-data settings. SAM3 first produces mushroom masks using macro-taxonomic prompts. DINOv3 then assigns labels through prototype matching in embedding space. A simple transformation of the DINOv3 feature space improves classification. Compared with class-specific prompting, the approach scales better and keeps segmentation cost low. Results span one-shot to few-hundred-shot regimes, giving, to the authors' knowledge, the first baseline for fine-grained semantic segmentation in low-data settings.

Catalog record

CREATORSebastian Cavada, Francesco Pelosin, Lapo Faggi
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE15897
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.22492

Browse the full library · About the Tower of Babel