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

arXiv: https://arxiv.org/abs/2605.22492 · HTML: https://mlsanity.com/paper/2605.22492/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 18 references that are in the map.
- First posted on arXiv 21 May 2026.

## Builds on (most cited first, 12 of 12)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features](https://mlsanity.com/paper/2502.14786/) (2025)
- [Optimal whitening and decorrelation](https://mlsanity.com/paper/1512.00809/) (2015)
- [Danish Fungi 2020 -- Not Just Another Image Recognition Dataset](https://mlsanity.com/paper/2103.10107/) (2021)
- [An Investigation into Whitening Loss for Self-supervised Learning](https://mlsanity.com/paper/2210.03586/) (2022)
- [Disentangling 3D Prototypical Networks For Few-Shot Concept Learning](https://mlsanity.com/paper/2011.03367/) (2020)
- [Learning Invariant Visual Representations for Planning with Joint-Embedding Predictive World Models](https://mlsanity.com/paper/2602.18639/) (2026)
- [FungiTastic: A multi-modal dataset and benchmark for image categorization](https://mlsanity.com/paper/2408.13632/) (2024)
- [Overview of PlantCLEF 2025: Multi-Species Plant Identification in Vegetation Quadrat Images](https://mlsanity.com/paper/2509.17602/) (2025)
- [Whitening Consistently Improves Self-Supervised Learning](https://mlsanity.com/paper/2408.07519/) (2024)

## Built on by (most cited first, 1 of 1)

- [Training-Free Generalized Few-Shot Segmentation through Open-Vocabulary Semantic Arbitration](https://mlsanity.com/paper/2606.09474/) (2026)

## Abstract (arXiv)

Fine-grained semantic segmentation requires both precise localization and discrimination between visually similar classes. In FungiTastic, this problem is further complicated by a long-tailed distribution and strong variation in image acquisition conditions. We propose a training-free two-stage framework that decouples segmentation from classification. SAM3 first produces class-agnostic mushroom masks using macro-taxonomic prompts, and DINOv3 then assigns fine-grained labels through prototype matching in the embedding space. To improve this stage, we apply a simple transformation of the DINOv3 feature space that improves prototype-based classification. Compared with class-specific prompting, our approach is more scalable and keeps the segmentation cost low. We report results from one-shot to few-hundred-shot regimes, providing, to the best of our knowledge, the first baseline for fine-grained semantic segmentation in low-data settings.