YOLOv3 and YOLOv3u#
Overview#
This document presents an overview of three closely related object detection models, namely YOLOv3, YOLOv3-Ultralytics, and YOLOv3u.
-
YOLOv3: This is the third version of the You Only Look Once (YOLO) object detection algorithm. Originally developed by Joseph Redmon, YOLOv3 improved on its predecessors by introducing predictions at three different scales.
-
YOLOv3-Ultralytics: This is the Ultralytics PyTorch implementation of YOLOv3, maintained in the ultralytics/yolov3 repository. It keeps the original anchor-based architecture while adding more pretrained models and easier customization.
-
YOLOv3u: This is an updated version of YOLOv3-Ultralytics that incorporates the anchor-free, objectness-free split head used in YOLOv8 models. YOLOv3u maintains the same backbone and neck architecture as YOLOv3 but with the updated detection head from YOLOv8.
Key Features#
-
YOLOv3: Introduced detection at three different scales, with 13x13, 26x26, and 52x52 output grids for a 416x416 input. This significantly improved detection accuracy for objects of different sizes. Additionally, YOLOv3 added multi-label predictions for each bounding box and a stronger feature extractor network, Darknet-53.
-
YOLOv3u: This updated model incorporates the anchor-free, objectness-free split head from YOLOv8. By eliminating the need for pre-defined anchor boxes and objectness scores, this detection head design can improve the model's ability to detect objects of varying sizes and shapes. This makes YOLOv3u more robust and accurate for object detection tasks.
Supported Tasks and Modes#
YOLOv3 is designed specifically for object detection tasks. Ultralytics supports three variants of YOLOv3: yolov3u, yolov3-tinyu and yolov3-sppu. The u in the name signifies that these utilize the anchor-free head of YOLOv8, unlike their original architecture which is anchor-based. The variants trade off accuracy and speed: YOLOv3-Tiny is the smallest and fastest, while YOLOv3-SPP adds spatial pyramid pooling to the standard YOLOv3 model.
All three models support the Training, Validation, Inference, and Export modes, covering every stage from development to model deployment.
| Model Type | Pretrained Weights | Tasks Supported | Training | Validation | Inference | Export |
|---|---|---|---|---|---|---|
| YOLOv3(u) | yolov3u.pt | Object Detection | ✅ | ✅ | ✅ | ✅ |
| YOLOv3-Tiny(u) | yolov3-tinyu.pt | Object Detection | ✅ | ✅ | ✅ | ✅ |
| YOLOv3-SPP(u) | yolov3-sppu.pt | Object Detection | ✅ | ✅ | ✅ | ✅ |
This table summarizes the supported tasks and modes for each YOLOv3 variant.
Usage Examples#
This example provides simple YOLOv3 training and inference examples.
PyTorch pretrained *.pt models as well as configuration *.yaml files can be passed to the YOLO() class to create a model instance in Python:
from ultralytics import YOLO
# Load a COCO-pretrained YOLOv3u model
model = YOLO("yolov3u.pt")
# Display model information (optional)
model.info()
# Train the model on the COCO8 example dataset for 100 epochs
results = model.train(data="coco8.yaml", epochs=100, imgsz=640)
# Run inference with the YOLOv3u model on the 'bus.jpg' image
results = model("path/to/bus.jpg")Citations and Acknowledgments#
If you use YOLOv3 in your research, please cite the original YOLO papers and the Ultralytics YOLOv3 repository:
@article{redmon2018yolov3,
title={YOLOv3: An Incremental Improvement},
author={Redmon, Joseph and Farhadi, Ali},
journal={arXiv preprint arXiv:1804.02767},
year={2018}
}Thank you to Joseph Redmon and Ali Farhadi for developing the original YOLOv3.
FAQ#
YOLOv3 is the third iteration of the YOLO (You Only Look Once) object detection algorithm developed by Joseph Redmon, known for its balance of accuracy and speed, predicting at three different scales (13x13, 26x26, and 52x52 grids for a 416x416 input). YOLOv3-Ultralytics is Ultralytics' adaptation of YOLOv3 that adds support for more pretrained models and facilitates easier model customization. YOLOv3u is an upgraded variant of YOLOv3-Ultralytics, integrating the anchor-free, objectness-free split head from YOLOv8, improving detection robustness and accuracy for various object sizes.
You can train a YOLOv3u model with either Python or the CLI:
Examplefrom ultralytics import YOLO # Load a COCO-pretrained YOLOv3u model model = YOLO("yolov3u.pt") # Train the model on the COCO8 example dataset for 100 epochs results = model.train(data="coco8.yaml", epochs=100, imgsz=640)For more comprehensive training options and guidelines, visit our Train mode documentation.
YOLOv3u improves upon YOLOv3 and YOLOv3-Ultralytics by incorporating the anchor-free, objectness-free split head used in YOLOv8 models. This upgrade eliminates the need for pre-defined anchor boxes and objectness scores, enhancing its capability to detect objects of varying sizes and shapes more precisely. This makes YOLOv3u a better choice for complex and diverse object detection tasks. For more information, refer to the Key Features section.
You can run inference with YOLOv3 models from either Python or the CLI:
Examplefrom ultralytics import YOLO # Load a COCO-pretrained YOLOv3u model model = YOLO("yolov3u.pt") # Run inference with the YOLOv3u model on the 'bus.jpg' image results = model("path/to/bus.jpg")Refer to the Inference mode documentation for more details on running YOLO models.
YOLOv3, YOLOv3-Tiny and YOLOv3-SPP primarily support object detection tasks. These models support the Training, Validation, Inference, and Export modes.
If you use YOLOv3 in your research, please cite the original YOLO papers and the Ultralytics YOLOv3 repository. Example BibTeX citation:
Citation@article{redmon2018yolov3, title={YOLOv3: An Incremental Improvement}, author={Redmon, Joseph and Farhadi, Ali}, journal={arXiv preprint arXiv:1804.02767}, year={2018} }For more citation details, refer to the Citations and Acknowledgments section.
Comments
Ask a question, share your results, or help someone else.