dinov2

PyTorch code and models for the DINOv2 self-supervised learning method, own data set and own adapted training.

Fork A fork of beneroth13/dinov2; the README may describe the upstream project.

Overview

This is the repository of Low-resource finetuning of foundation models beats state-of-the-art in histopathology which was accepted at ISBI 2024. It is a slightly adapted version of the original DINOv2, GitHub repository.

We propose finetuning a DINOv2 ViT-S, which yields at least equal performance compared to CTransPath and RetCCL but in a fraction of domain specific training time. Performance is measured on three datasets: TCGA & CPTAC (WSI-level classification) and NCT-CRC (patch-level classification).

Performance over time of finetuning a ViT-s with DINOv2: a) on NCT-CRC and evaluating on the external NCT- CRC testset on patch-level classification and b) on TCGA and testing on TCGA (5-fold cross-validation) and CPTAC (external testset) on WSI-level classification.

For the finetuning process, we utilized histopathological data from two primary datasets:

From the project’s README on GitHub.

At a glance

Repository	Kentucky-Open-Science/dinov2
Research area	Forks of other projects
Primary language	None detected
Languages	Jupyter Notebook 67.9%, Python 32.1%
License	Apache-2.0
Stars / forks	0 / 0
Open issues and pull requests	4
Created	2024-09-23
Last push	2026-03-12
Default branch	`main`
Forked from	beneroth13/dinov2 — PyTorch code and models for the DINOv2 self-supervised learning method, own data set and own adapted training.

What the README covers

Top contributors

@beneroth13 (106 commits)
@patricklabatut (22 commits)
@ValentinKoch (3 commits)
@leo-gan (2 commits)
@goggle (1 commit)
@Aryanutkarsh (1 commit)
@qasfb (1 commit)

Get the code

git clone https://github.com/Kentucky-Open-Science/dinov2.git