DINOv3 Meets YOLO26 for Weed Detection in Vegetable Crops

TL;DR

This study introduces a novel weed detection model combining DINOv3 and YOLO26, leveraging self-supervised learning and heterogeneous datasets to improve accuracy and generalization in vegetable crop weeding.

Contribution

It presents a new integrated model using a fine-tuned DINOv3 backbone within YOLO26, enhancing detection performance and cross-domain robustness for precision agriculture.

Findings

Achieved up to +5.4% mAP50 on in-domain data

Improved cross-domain mAP50 by +14.0% and +11.9%

Maintains real-time detection at ~28.5 fps

Abstract

Developing robust models for precision vegetable weeding is currently constrained by the scarcity of large-scale, annotated weed-crop datasets. To address this limitation, this study proposes a foundational crop-weed detection model by integrating heterogeneous datasets and leveraging self-supervised learning. A total of 618,642 crop-weed images were initially collected and subsequently refined to 199,388 filtered images for fine-tuning a DINOv3 vision transformer (ViT-small) through a sequential curation strategy. The fine-tuned DINOv3 backbone was then integrated into YOLO26, serving either as a primary backbone or part of a dual-backbone architecture. A feature alignment loss was introduced in the dual backbone framework to enhance feature fusion with minimal computational overhead. Experimental results show that the proposed DINOv3-finetuned ViT-small-based YOLO26-large achieved up to a +5.4% mAP50 gain on in-domain images collected in the 2025 season. Moreover, it demonstrated strong cross-domain generalization with mAP50 improvements of +14.0% on the 2021-2023 season dataset and +11.9% on the 2024 season dataset, compared to the standard YOLO26-large. Although the DINOv3-YOLO26-large model has 45.6% more parameters and a 2.9x increase in inference latency, it maintains real-time performance at ~28.5 frames per second (fps). The curated dataset and software programs developed in this study will be made publicly available.