Evaluating Vision Foundation Models for Pixel and Object Classification in Microscopy

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Teuber, Carolin, Archit, Anwai, Boothe, Tobias, Ditte, Peter, Rink, Jochen, Pape, Constantin
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908902672564224
author Teuber, Carolin
Archit, Anwai
Boothe, Tobias
Ditte, Peter
Rink, Jochen
Pape, Constantin
author_facet Teuber, Carolin
Archit, Anwai
Boothe, Tobias
Ditte, Peter
Rink, Jochen
Pape, Constantin
contents Deep learning underlies most modern approaches and tools in computer vision, including biomedical imaging. However, for interactive semantic segmentation (often called pixel classification in this context) and interactive object-level classification (object classification), feature-based shallow learning remains widely used. This is due to the diversity of data in this domain, the lack of large pretraining datasets, and the need for computational and label efficiency. In contrast, state-of-the-art tools for many other vision tasks in microscopy - most notably cellular instance segmentation - already rely on deep learning and have recently benefited substantially from vision foundation models (VFMs), particularly SAM. Here, we investigate whether VFMs can also improve pixel and object classification compared to current approaches. To this end, we evaluate several VFMs, including general-purpose models (SAM, SAM2, DINOv3) and domain-specific ones ($μ$SAM, PathoSAM), in combination with shallow learning and attentive probing on five diverse and challenging datasets. Our results demonstrate consistent improvements over hand-crafted features and provide a clear pathway toward practical improvements. Furthermore, our study establishes a benchmark for VFMs in microscopy and informs future developments in this area.
format Preprint
id arxiv_https___arxiv_org_abs_2603_19802
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Evaluating Vision Foundation Models for Pixel and Object Classification in Microscopy
Teuber, Carolin
Archit, Anwai
Boothe, Tobias
Ditte, Peter
Rink, Jochen
Pape, Constantin
Computer Vision and Pattern Recognition
Deep learning underlies most modern approaches and tools in computer vision, including biomedical imaging. However, for interactive semantic segmentation (often called pixel classification in this context) and interactive object-level classification (object classification), feature-based shallow learning remains widely used. This is due to the diversity of data in this domain, the lack of large pretraining datasets, and the need for computational and label efficiency. In contrast, state-of-the-art tools for many other vision tasks in microscopy - most notably cellular instance segmentation - already rely on deep learning and have recently benefited substantially from vision foundation models (VFMs), particularly SAM. Here, we investigate whether VFMs can also improve pixel and object classification compared to current approaches. To this end, we evaluate several VFMs, including general-purpose models (SAM, SAM2, DINOv3) and domain-specific ones ($μ$SAM, PathoSAM), in combination with shallow learning and attentive probing on five diverse and challenging datasets. Our results demonstrate consistent improvements over hand-crafted features and provide a clear pathway toward practical improvements. Furthermore, our study establishes a benchmark for VFMs in microscopy and informs future developments in this area.
title Evaluating Vision Foundation Models for Pixel and Object Classification in Microscopy
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.19802