Ultralytics YOLOv8 Classification

Ultralytics YOLOv8 is the latest version of the acclaimed real-time object detection and image segmentation model. The YOLO (You Only Look Once) v8 model is built on cutting-edge advancements in deep learning and computer vision, offering unparalleled performance in terms of speed and accuracy.

Intended Use

The YOLOv8 image classification model is designed to detect 1000 pre-defined classes in images in real-time.

Image classification is the simplest of the three tasks and involves classifying an entire image into one of a set of predefined classes. Different from YOLO's Segment and Object detection models which are trained on COCO datasets, the image classification models are trained on ImageNet dataset. YOLOv8 comes bundled with image classification models pre-trained on the ImageNet dataset with an image resolution of 224.

Performance

Compared to the previous YOLOv5 model, YOLOv8 comes with many features built for developers, from an easy-to-use CLI to a well-structured python package in addition to a higher rate of accuracy.

The YOLOv8 model has demonstrated state-of-the-art performance on ImageNet classification tasks. When evaluated side by side on datasets for task-specific domains, YOLOv8 scored substantially better than YOLOv5. You can learn more about usage, performance, and key features here.

Limitations

The YOLO algorithm has a few limitations such as struggling with small objects and the inability to perform fine-grained classification, however it still has simple architecture and requires minimal training data, making it easy to implement and adapt to new tasks.

Citation

Authors: Glenn Jocher and Ayush Chaurasia and Jing Qiu

Title: Ultralytics YOLOv8

Version: 8.0.0

Year: 2023

URL: https://github.com/ultralytics/ultralytics

Orcid: 0000-0001-5950-6979, 0000-0002-7603-6750, 0000-0003-3783-7069

License: AGPL-3.0