AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

DINOv3

Oriane Siméoni, Huy Vo, Maximilian Seitzer, Federico Baldassarre, Maxime Oquab, Cijo Jose, Vasil Khalidov, Marc Szafraniec, Seung Eun Yi, Michaël Ramamonjisoa, Francisco Massa, Daniel Haziza and 14 more

NeurIPS 2026 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Invariance-based self-supervised learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Technical report on DINOv3, a self-supervised visual foundation model aimed at learning general-purpose representations from diverse image sources (natural to aerial) with a single algorithm.

Problem

Self-supervised learning should remove the need for manual annotation and scale to massive datasets and larger architectures, learning visual representations across diverse domains with one algorithm.

Approach

Scaling a self-supervised vision training recipe (DINO lineage) to learn general visual representations from diverse sources using a single algorithm; details beyond the abstract are not available.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use combined curated data mixture (clustering + retrieval + public datasets) instead of raw data (ablation)DINOv3 (200k iterations) · iNaturalist 2021accuracy: 70.1 -> 87.0Improved
use data mixture instead of retrieval-only curation (ablation)DINOv3 (200k iterations) · Paris retrievalaccuracy: 82.7 -> 85.9Improved
use clustering-only curation instead of raw data (ablation)DINOv3 (200k iterations) · IN1k k-NNaccuracy: 80.1 -> 79.4Hurt
add Gram anchoring refinement (ablation)DINOv3 ViT-7B · ADE20kmIoUImproved
use high-resolution Gram teacher features (L_HRef vs L_Ref) (ablation)DINOv3 ViT-7B · ADE20kmIoU: +2 mIoUImproved
use a much later Gram teacher (1M iterations) instead of 100k/200k (ablation)DINOv3 · dense tasksmIoUHurt
train longer without Gram anchoring (ablation)DINOv3 ViT-7B · Pascal VOC segmentationmIoUHurt
add dedicated layer norm for local and global crop outputs (ablation)DINOv3 · ImageNet kNN / ADE20k / NYUv2kNN acc / mIoU / RMSE: +0.2 acc, +1 mIoU, -0.02 RMSEImproved
apply high-resolution adaptation (ablation)DINOv3 ViT-7B · ObjectNet at low resolutionaccuracyMixed
omit Gram anchoring in high-resolution adaptation (ablation)DINOv3 ViT-7B · dense predictionmIoUHurt
scale to ViT-7B with SSL and Gram anchoring vs DINOv2DINOv3 · KITTI depthRMSE: 2.624 -> 2.346Improved
use SSL-only DINOv3 vs weakly-supervised PE-core on OCR tasksDINOv3 · GTSRBaccuracy: 94.8 -> 87.5Hurt

Limitations

Experimental setup

Builds on: DINO DINOv2

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy