RF-DETR

RF-DETR is a real-time object detection transformer-based architecture designed to transfer well to both a wide variety of domains and to datasets big and small.

RF-DETR is the first real-time model to exceed 60 AP on the Microsoft COCO benchmark alongside competitive performance at base sizes. It also achieves state-of-the-art performance on RF100-VL, an object detection benchmark that measures model domain adaptability to real world problems. RF-DETR is comparable speed to current real-time objection models.

RF-DETR is developed for projects that need a model that can run high speeds with a high degree of accuracy, and often on limited compute (like on the edge or low latency).

The model comes in two variants:

• RF-DETR Base, which has 29M parameters, and;
• RF-DETR Large, which has 129M parameters.

The model is available under an Apache 2.0 license, and the accompanying paper "RF-DETR: Neural Architecture Search for Real-Time Detection Transformers" is available on Arxiv.

Running RF-DETR

You can run RF-DETR on COCO using the rfdetr  Python package:

import io
import requests
import supervision as sv
from PIL import Image
from rfdetr import RFDETRBase

model = RFDETRBase()

url = "https://media.roboflow.com/notebooks/examples/dog-2.jpeg"

image = Image.open(io.BytesIO(requests.get(url).content))
detections = model.predict(image, threshold=0.5)

annotated_image = image.copy()
annotated_image = sv.BoxAnnotator().annotate(annotated_image, detections)
annotated_image = sv.LabelAnnotator().annotate(annotated_image, detections)

sv.plot_image(annotated_image)

Here is an example showing the model run on a random selection of images from the Microsoft COCO validation set: