RF-DETR

Detects objects in real time, and fine-tunes on a small set of local annotations.

Field
Cross-domain
Method
Deep learning
Data it takes
imagery
License
Apache-2.0

RF-DETR is a detection transformer on a DINOv2 backbone, built to be fine-tuned on a modest annotation set.

It is the permissively licensed alternative to YOLO. YOLO is AGPL, which reaches any work that redistributes or serves the model, and that makes it awkward for results the lab hands to another group.

Use it where the classes are fixed and worth annotating. Where they are not, Grounding DINO detects from a text prompt with no annotation at all.

Upstream project → · Paper →

Catalog entry last checked 2026-09-08. All models →

Run RF-DETR on your data

The lab runs this on its own compute, in a container, with the inputs and parameters recorded alongside the result. Initial scoping conversations are free.

Contact the Lab