YOLO11 is a pc imaginative and prescient mannequin structure developed by Ultralytics, the creators of the YOLOv5 and YOLOv8 fashions. YOLO11 helps object detection, segmentation, classification, keypoint detection, and oriented bounding field (OBB) detection.
On this information, we’re going to talk about what YOLO11 is, the way it performs, and how one can prepare and deploy YOLO11 fashions to your personal {hardware}.
With out additional ado, let’s get began!
What’s YOLO11?
YOLO11 is a collection of pc imaginative and prescient fashions developed by Ultralytics. As of the launch of YOLO11, the mannequin is probably the most correct of all Ultralytics’ fashions.
The YOLO11x mannequin, the biggest within the collection, reportedly achieves a 54.7% mAP rating when evaluated towards the Microsoft COCO benchmark. The smallest mannequin, YOLO11n reportedly achieves a 39.5% mAP rating when evaluated towards the identical dataset. See how YOLO11 compares to different object detection fashions within the object detection mannequin leaderboard.
YOLO11 has assist for a similar job varieties as YOLOv8, that are:
- Object detection
- Classification
- Picture segmentation
- Keypoint detection
- Oriented Bounding Field (OBB)
Getting Began with YOLO11
To get began making use of YOLO11 to your personal use case, try our information on the right way to prepare YOLOv8 on customized dataset.
To see what others are doing with YOLO11, browse Roboflow Universe for different YOLO11 fashions, datasets, and inspiration.
Deploy YOLO11 Fashions to the Edge
Along with utilizing the Roboflow hosted API for deployment, you should use Roboflow Inference, an open supply inference answer. Inference works with CPU and GPU, providing you with quick entry to a variety of gadgets, from the NVIDIA Jetson (i.e. Jetson Nano or Orin) to ARM CPU gadgets.
With Roboflow Inference, you’ll be able to self-host and deploy your mannequin on-device.
Step #1: Add Mannequin Weights to Roboflow
Upon getting completed coaching a YOLOv8 mannequin, you should have a set of skilled weights prepared to be used with a hosted API endpoint. You may add your mannequin weights to Roboflow Deploy with the deploy() operate within the Roboflow pip package deal to make use of your skilled weights within the cloud.
To add mannequin weights, first create a brand new undertaking on Roboflow, add your dataset, and create a undertaking model. Take a look at our full information on the right way to create and arrange a undertaking in Roboflow. Then, write a Python script with the next code:
import roboflow
roboflow.login()
rf = roboflow.Roboflow()
undertaking = rf.workspace().undertaking(PROJECT_ID)
undertaking.model(DATASET_VERSION).deploy(model_type="yolo11", model_path=f"{HOME}/runs/detect/prepare/")Change PROJECT_ID with the ID of your undertaking and DATASET_VERSION with the model quantity related together with your undertaking. Discover ways to discover your undertaking ID and dataset model quantity.
Shortly after working the above code, your mannequin will likely be accessible to be used within the Deploy web page in your Roboflow undertaking dashboard.
Step #2: Set up Inference
You may deploy purposes utilizing the Inference Docker containers or the pip package deal. Let’s use the pip package deal. First run:
pip set up inferenceStep #3: Run Inference on an Picture
Then, create a brand new Python file and add the next code:
from inference import get_model
import supervision as sv
import cv2 image_file = "picture.jpeg"
picture = cv2.imread(image_file)
mannequin = get_model(model_id="model-id") # run inference on our chosen picture, picture could be a url, a numpy array, a PIL picture, and so on.
outcomes = mannequin.infer(picture)[0] # load the outcomes into the supervision Detections api
detections = sv.Detections.from_inference(outcomes) # create supervision annotators
bounding_box_annotator = sv.BoundingBoxAnnotator()
label_annotator = sv.LabelAnnotator() # annotate the picture with our inference outcomes
annotated_image = bounding_box_annotator.annotate(
scene=picture, detections=detections)
annotated_image = label_annotator.annotate(
scene=annotated_image, detections=detections) # show the picture
sv.plot_image(annotated_image)Above, set your Roboflow workspace ID, mannequin ID, and API key, if you wish to use a customized mannequin you’ve got skilled in your workspace.
Additionally, set the URL of a picture on which you wish to run inference. This could be a native file.
Right here is an instance of a picture working by means of the mannequin:
The mannequin efficiently detected an individual within the picture, indicated by the purple bounding field on the picture.
You may as well run inference on a video stream. To study extra about working your mannequin on video streams – from RTSP to webcam feeds – consult with the Inference video information.
YOLO11 FAQs
Does YOLO11 have a broadcast paper?
YOLO11 doesn’t have a broadcast or pre-print tutorial paper.
Below what license is YOLO11 lined?
YOLO11 is roofed underneath an AGPL-3.zero license. For those who deploy YOLO11 fashions with Roboflow, you mechanically get a industrial license to make use of the mannequin.
The place is the YOLO11 supply code?
You could find the YOLO11 supply code within the Ultralytics GitHub repository.
What courses can the bottom YOLO11 weights establish?
The bottom YOLO11 weights that had been launched with the mannequin had been skilled on the Microsoft COCO dataset. You may see a full record of the COCO courses right here.
Conclusion
Introduced and launched in September 2024, YOLO11 is the newest collection of pc imaginative and prescient fashions developed by Ultralytics. The mannequin structure has assist for all the identical imaginative and prescient duties as YOLOv8, whereas providing improved accuracy when evaluated towards the COCO dataset benchmark.
On this information, we walked by means of the fundamentals of YOLO11: what the mannequin is, what you are able to do with it, and the right way to deploy the mannequin by yourself machine. To study extra about coaching YOLO11 fashions, consult with our YOLO11 coaching information.
