# Upsampling DINOv2 features for unsupervised vision tasks and weakly supervised materials segmentation

Source: [Upsampling DINOv2 features for unsupervised vision tasks and weakly supervised materials segmentation](<https://www.sota2.com/research/paper/68fa4b4e1e209131ce944f34>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T08:15:58.165Z

Authors: Ronan Docherty, Antonis Vamvakeros, Samuel J. Cooper

Published: 2024-10-20T13:01:53.000Z

## Abstract

The features of self-supervised vision transformers (ViTs) contain strong semantic and positional information relevant to downstream tasks like object localization and segmentation. Recent works combine these features with traditional methods like clustering, graph partitioning or region correlations to achieve impressive baselines without finetuning or training additional networks. We leverage upsampled features from ViT networks (e.g DINOv2) in two workflows: in a clustering based approach for object localization and segmentation, and paired with standard classifiers in weakly supervised materials segmentation. Both show strong performance on benchmarks, especially in weakly supervised segmentation where the ViT features capture complex relationships inaccessible to classical approaches. We expect the flexibility and generalizability of these features will both speed up and strengthen materials characterization, from segmentation to property-prediction.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2410.19836>)

[PDF](<https://arxiv.org/pdf/2410.19836.pdf>)
