description: Learn how to install the Ultralytics package in developer mode and build/serve locally using MkDocs. Deploy your project to your host easily.
---
# Ultralytics Docs
# Ultralytics Docs
Ultralytics Docs are deployed to [https://docs.ultralytics.com](https://docs.ultralytics.com).
Ultralytics Docs are deployed to [https://docs.ultralytics.com](https://docs.ultralytics.com).
description: Learn how torchvision organizes classification image datasets. Use this code to create and train models. CLI and Python code shown.
---
---
# Image Classification Datasets Overview
# Image Classification Datasets Overview
@ -77,6 +78,7 @@ cifar-10-/
In this example, the `train` directory contains subdirectories for each class in the dataset, and each class subdirectory contains all the images for that class. The `test` directory has a similar structure. The `root` directory also contains other files that are part of the CIFAR10 dataset.
In this example, the `train` directory contains subdirectories for each class in the dataset, and each class subdirectory contains all the images for that class. The `test` directory has a similar structure. The `root` directory also contains other files that are part of the CIFAR10 dataset.
## Usage
## Usage
!!! example ""
!!! example ""
=== "Python"
=== "Python"
@ -98,4 +100,5 @@ In this example, the `train` directory contains subdirectories for each class in
description: Learn about the COCO dataset, designed to encourage research on object detection, segmentation, and captioning with standardized evaluation metrics.
description: Learn about supported dataset formats for training YOLO detection models, including Ultralytics YOLO and COCO, in this Object Detection Datasets Overview.
---
---
# Object Detection Datasets Overview
# Object Detection Datasets Overview
@ -20,6 +21,7 @@ The dataset format used for training YOLO detection models is as follows:
- Object width and height: The width and height of the object, normalized to be between 0 and 1.
- Object width and height: The width and height of the object, normalized to be between 0 and 1.
The format for a single row in the detection dataset file is as follows:
The format for a single row in the detection dataset file is as follows:
```
```
<object-class><x><y><width><height>
<object-class><x><y><width><height>
```
```
@ -55,6 +57,7 @@ The `names` field is a list of the names of the object classes. The order of the
NOTE: Either `nc` or `names` must be defined. Defining both are not mandatory
NOTE: Either `nc` or `names` must be defined. Defining both are not mandatory
Alternatively, you can directly define class names like this:
Alternatively, you can directly define class names like this:
description: Ultralytics provides support for various datasets to facilitate multiple computer vision tasks. Check out our list of main datasets and their summaries.
---
---
# Datasets Overview
# Datasets Overview
@ -10,44 +11,44 @@ Ultralytics provides support for various datasets to facilitate computer vision
Bounding box object detection is a computer vision technique that involves detecting and localizing objects in an image by drawing a bounding box around each object.
Bounding box object detection is a computer vision technique that involves detecting and localizing objects in an image by drawing a bounding box around each object.
* [Argoverse](detect/argoverse.md): A dataset containing 3D tracking and motion forecasting data from urban environments with rich annotations.
* [Argoverse](detect/argoverse.md): A dataset containing 3D tracking and motion forecasting data from urban environments with rich annotations.
* [COCO](detect/coco.md): A large-scale dataset designed for object detection, segmentation, and captioning with over 200K labeled images.
* [COCO](detect/coco.md): A large-scale dataset designed for object detection, segmentation, and captioning with over 200K labeled images.
* [COCO8](detect/coco8.md): Contains the first 4 images from COCO train and COCO val, suitable for quick tests.
* [COCO8](detect/coco8.md): Contains the first 4 images from COCO train and COCO val, suitable for quick tests.
* [Global Wheat 2020](detect/globalwheat2020.md): A dataset of wheat head images collected from around the world for object detection and localization tasks.
* [Global Wheat 2020](detect/globalwheat2020.md): A dataset of wheat head images collected from around the world for object detection and localization tasks.
* [Objects365](detect/objects365.md): A high-quality, large-scale dataset for object detection with 365 object categories and over 600K annotated images.
* [Objects365](detect/objects365.md): A high-quality, large-scale dataset for object detection with 365 object categories and over 600K annotated images.
* [SKU-110K](detect/sku-110k.md): A dataset featuring dense object detection in retail environments with over 11K images and 1.7 million bounding boxes.
* [SKU-110K](detect/sku-110k.md): A dataset featuring dense object detection in retail environments with over 11K images and 1.7 million bounding boxes.
* [VisDrone](detect/visdrone.md): A dataset containing object detection and multi-object tracking data from drone-captured imagery with over 10K images and video sequences.
* [VisDrone](detect/visdrone.md): A dataset containing object detection and multi-object tracking data from drone-captured imagery with over 10K images and video sequences.
* [VOC](detect/voc.md): The Pascal Visual Object Classes (VOC) dataset for object detection and segmentation with 20 object classes and over 11K images.
* [VOC](detect/voc.md): The Pascal Visual Object Classes (VOC) dataset for object detection and segmentation with 20 object classes and over 11K images.
* [xView](detect/xview.md): A dataset for object detection in overhead imagery with 60 object categories and over 1 million annotated objects.
* [xView](detect/xview.md): A dataset for object detection in overhead imagery with 60 object categories and over 1 million annotated objects.
Instance segmentation is a computer vision technique that involves identifying and localizing objects in an image at the pixel level.
Instance segmentation is a computer vision technique that involves identifying and localizing objects in an image at the pixel level.
* [COCO](segment/coco.md): A large-scale dataset designed for object detection, segmentation, and captioning tasks with over 200K labeled images.
* [COCO](segment/coco.md): A large-scale dataset designed for object detection, segmentation, and captioning tasks with over 200K labeled images.
* [COCO8-seg](segment/coco8-seg.md): A smaller dataset for instance segmentation tasks, containing a subset of 8 COCO images with segmentation annotations.
* [COCO8-seg](segment/coco8-seg.md): A smaller dataset for instance segmentation tasks, containing a subset of 8 COCO images with segmentation annotations.
## [Pose Estimation](pose/index.md)
## [Pose Estimation](pose/index.md)
Pose estimation is a technique used to determine the pose of the object relative to the camera or the world coordinate system.
Pose estimation is a technique used to determine the pose of the object relative to the camera or the world coordinate system.
* [COCO](pose/coco.md): A large-scale dataset with human pose annotations designed for pose estimation tasks.
* [COCO](pose/coco.md): A large-scale dataset with human pose annotations designed for pose estimation tasks.
* [COCO8-pose](pose/coco8-pose.md): A smaller dataset for pose estimation tasks, containing a subset of 8 COCO images with human pose annotations.
* [COCO8-pose](pose/coco8-pose.md): A smaller dataset for pose estimation tasks, containing a subset of 8 COCO images with human pose annotations.
## [Classification](classify/index.md)
## [Classification](classify/index.md)
Image classification is a computer vision task that involves categorizing an image into one or more predefined classes or categories based on its visual content.
Image classification is a computer vision task that involves categorizing an image into one or more predefined classes or categories based on its visual content.
* [Caltech 101](classify/caltech101.md): A dataset containing images of 101 object categories for image classification tasks.
* [Caltech 101](classify/caltech101.md): A dataset containing images of 101 object categories for image classification tasks.
* [Caltech 256](classify/caltech256.md): An extended version of Caltech 101 with 256 object categories and more challenging images.
* [Caltech 256](classify/caltech256.md): An extended version of Caltech 101 with 256 object categories and more challenging images.
* [CIFAR-10](classify/cifar10.md): A dataset of 60K 32x32 color images in 10 classes, with 6K images per class.
* [CIFAR-10](classify/cifar10.md): A dataset of 60K 32x32 color images in 10 classes, with 6K images per class.
* [CIFAR-100](classify/cifar100.md): An extended version of CIFAR-10 with 100 object categories and 600 images per class.
* [CIFAR-100](classify/cifar100.md): An extended version of CIFAR-10 with 100 object categories and 600 images per class.
* [Fashion-MNIST](classify/fashion-mnist.md): A dataset consisting of 70,000 grayscale images of 10 fashion categories for image classification tasks.
* [Fashion-MNIST](classify/fashion-mnist.md): A dataset consisting of 70,000 grayscale images of 10 fashion categories for image classification tasks.
* [ImageNet](classify/imagenet.md): A large-scale dataset for object detection and image classification with over 14 million images and 20,000 categories.
* [ImageNet](classify/imagenet.md): A large-scale dataset for object detection and image classification with over 14 million images and 20,000 categories.
* [ImageNet-10](classify/imagenet10.md): A smaller subset of ImageNet with 10 categories for faster experimentation and testing.
* [ImageNet-10](classify/imagenet10.md): A smaller subset of ImageNet with 10 categories for faster experimentation and testing.
* [Imagenette](classify/imagenette.md): A smaller subset of ImageNet that contains 10 easily distinguishable classes for quicker training and testing.
* [Imagenette](classify/imagenette.md): A smaller subset of ImageNet that contains 10 easily distinguishable classes for quicker training and testing.
* [Imagewoof](classify/imagewoof.md): A more challenging subset of ImageNet containing 10 dog breed categories for image classification tasks.
* [Imagewoof](classify/imagewoof.md): A more challenging subset of ImageNet containing 10 dog breed categories for image classification tasks.
* [MNIST](classify/mnist.md): A dataset of 70,000 grayscale images of handwritten digits for image classification tasks.
* [MNIST](classify/mnist.md): A dataset of 70,000 grayscale images of handwritten digits for image classification tasks.
In this format, `<class-index>` is the index of the class for the object,`<x><y><width><height>` are coordinates of boudning box, and `<px1> <py1> <px2> <py2> <pxn> <pyn>` are the pixel coordinates of the keypoints. The coordinates are separated by spaces.
In this format, `<class-index>` is the index of the class for the object,`<x><y><width><height>` are coordinates of boudning box, and `<px1> <py1> <px2> <py2> ... <pxn> <pyn>` are the pixel coordinates of the keypoints. The coordinates are separated by spaces.
** Dataset file format **
** Dataset file format **
@ -62,6 +63,7 @@ The `names` field is a list of the names of the object classes. The order of the
NOTE: Either `nc` or `names` must be defined. Defining both are not mandatory
NOTE: Either `nc` or `names` must be defined. Defining both are not mandatory
Alternatively, you can directly define class names like this:
Alternatively, you can directly define class names like this:
description: Learn about the Ultralytics YOLO dataset format for segmentation models. Use YAML to train Detection Models. Convert COCO to YOLO format using Python.
---
---
# Instance Segmentation Datasets Overview
# Instance Segmentation Datasets Overview
@ -32,6 +33,7 @@ Here is an example of the YOLO dataset format for a single image with two object
description: Discover the datasets compatible with Multi-Object Detector. Train your trackers and make your detections more efficient with Ultralytics' YOLO.
---
---
# Multi-object Tracking Datasets Overview
# Multi-object Tracking Datasets Overview
@ -26,4 +27,3 @@ Support for training trackers alone is coming soon
```bash
```bash
yolo track model=yolov8n.pt source="https://youtu.be/Zgi9g1ksQHc" conf=0.3, iou=0.5 show
yolo track model=yolov8n.pt source="https://youtu.be/Zgi9g1ksQHc" conf=0.3, iou=0.5 show
description: Individual Contributor License Agreement. Settle Intellectual Property issues for Contributions made to anything open source released by Ultralytics.
description: Read the Ultralytics Contributor Covenant Code of Conduct. Learn ways to create a welcoming community & consequences for inappropriate conduct.
---
---
# Ultralytics Contributor Covenant Code of Conduct
# Ultralytics Contributor Covenant Code of Conduct
description: Learn how to contribute to Ultralytics Open-Source YOLO Repositories with contributions guidelines, pull requests requirements, and GitHub CI tests.
---
---
# Contributing to Ultralytics Open-Source YOLO Repositories
# Contributing to Ultralytics Open-Source YOLO Repositories
description: Get comprehensive resources for Ultralytics YOLO repositories. Find guides, FAQs, MRE creation, CLA & more. Join the supportive community now!
---
---
Welcome to the Ultralytics Help page! We are committed to providing you with comprehensive resources to make your experience with Ultralytics YOLO repositories as smooth and enjoyable as possible. On this page, you'll find essential links to guides and documents that will help you navigate through common tasks and address any questions you might have while using our repositories.
Welcome to the Ultralytics Help page! We are committed to providing you with comprehensive resources to make your experience with Ultralytics YOLO repositories as smooth and enjoyable as possible. On this page, you'll find essential links to guides and documents that will help you navigate through common tasks and address any questions you might have while using our repositories.
description: Learn how to create a Minimum Reproducible Example (MRE) for Ultralytics YOLO bug reports to help maintainers and contributors understand your issue better.
---
---
# Creating a Minimum Reproducible Example for Bug Reports in Ultralytics YOLO Repositories
# Creating a Minimum Reproducible Example for Bug Reports in Ultralytics YOLO Repositories
description: Run YOLO models on your Android device for real-time object detection with Ultralytics Android App. Utilizes TensorFlow Lite and hardware delegates.
---
---
# Ultralytics Android App: Real-time Object Detection with YOLO Models
# Ultralytics Android App: Real-time Object Detection with YOLO Models
description: Get started with the Ultralytics iOS app and run YOLO models in real-time for object detection on your iPhone or iPad with the Apple Neural Engine.
---
---
# Ultralytics iOS App: Real-time Object Detection with YOLO Models
# Ultralytics iOS App: Real-time Object Detection with YOLO Models
@ -33,7 +34,6 @@ By combining quantized YOLO models with the Apple Neural Engine, the Ultralytics
description: 'Ultralytics HUB: Train & deploy YOLO models from one spot! Use drag-and-drop interface with templates & pre-training models. Check quickstart, datasets, and more.'
---
---
# Ultralytics HUB
# Ultralytics HUB
@ -20,7 +21,6 @@ comments: true
launch [Ultralytics HUB](https://bit.ly/ultralytics_hub), a new web tool for training and deploying all your YOLOv5 and YOLOv8 🚀
launch [Ultralytics HUB](https://bit.ly/ultralytics_hub), a new web tool for training and deploying all your YOLOv5 and YOLOv8 🚀
models from one spot!
models from one spot!
## Introduction
## Introduction
HUB is designed to be user-friendly and intuitive, with a drag-and-drop interface that allows users to
HUB is designed to be user-friendly and intuitive, with a drag-and-drop interface that allows users to
This page is currently under construction!️ 👷Please check back later for updates. 😃🔜
This page is currently under construction!️ 👷Please check back later for updates. 😃🔜
# YOLO Inference API
# YOLO Inference API
The YOLO Inference API allows you to access the YOLOv8 object detection capabilities via a RESTful API. This enables you to run object detection on images without the need to install and set up the YOLOv8 environment locally.
The YOLO Inference API allows you to access the YOLOv8 object detection capabilities via a RESTful API. This enables you to run object detection on images without the need to install and set up the YOLOv8 environment locally.
@ -45,7 +44,6 @@ print(response.json())
In this example, replace `API_KEY` with your actual API key, `MODEL_ID` with the desired model ID, and `path/to/image.jpg` with the path to the image you want to analyze.
In this example, replace `API_KEY` with your actual API key, `MODEL_ID` with the desired model ID, and `path/to/image.jpg` with the path to the image you want to analyze.
## Example Usage with CLI
## Example Usage with CLI
You can use the YOLO Inference API with the command-line interface (CLI) by utilizing the `curl` command. Replace `API_KEY` with your actual API key, `MODEL_ID` with the desired model ID, and `image.jpg` with the path to the image you want to analyze:
You can use the YOLO Inference API with the command-line interface (CLI) by utilizing the `curl` command. Replace `API_KEY` with your actual API key, `MODEL_ID` with the desired model ID, and `image.jpg` with the path to the image you want to analyze:
@ -334,7 +332,6 @@ YOLO segmentation models, such as `yolov8n-seg.pt`, can return JSON responses fr
}
}
```
```
### Pose Model Format
### Pose Model Format
YOLO pose models, such as `yolov8n-pose.pt`, can return JSON responses from local inference, CLI API inference, and Python API inference. All of these methods produce the same JSON response format.
YOLO pose models, such as `yolov8n-pose.pt`, can return JSON responses from local inference, CLI API inference, and Python API inference. All of these methods produce the same JSON response format.
description: Explore Ultralytics YOLOv8, a cutting-edge real-time object detection and image segmentation model for various applications and hardware platforms.
description: Learn about the supported models and architectures, such as YOLOv3, YOLOv5, and YOLOv8, and how to contribute your own model to Ultralytics.
description: Detect objects faster and more accurately using Ultralytics YOLOv5u. Find pre-trained models for each task, including Inference, Validation and Training.
description: Benchmark mode compares speed and accuracy of various YOLOv8 export formats like ONNX or OpenVINO. Optimize formats for speed or accuracy.
description: 'Export mode: Create a deployment-ready YOLOv8 model by converting it to various formats. Export to ONNX or OpenVINO for up to 3x CPU speedup.'
| `conf (bool)` | Whether to plot the detection confidence score. |
| `conf (bool)` | Whether to plot the detection confidence score. |
@ -234,7 +237,6 @@ masks, classification logits, etc.) found in the results object
| `masks (bool)` | Whether to plot the masks. |
| `masks (bool)` | Whether to plot the masks. |
| `probs (bool)` | Whether to plot classification probability. |
| `probs (bool)` | Whether to plot classification probability. |
## Streaming Source `for`-loop
## Streaming Source `for`-loop
Here's a Python script using OpenCV (cv2) and YOLOv8 to run inference on video frames. This script assumes you have already installed the necessary packages (opencv-python and ultralytics).
Here's a Python script using OpenCV (cv2) and YOLOv8 to run inference on video frames. This script assumes you have already installed the necessary packages (opencv-python and ultralytics).
description: Install and use YOLOv8 via CLI or Python. Run single-line commands or integrate with Python projects for object detection, segmentation, and classification.
---
---
## Install
## Install
@ -32,13 +33,11 @@ See the `ultralytics` [requirements.txt](https://github.com/ultralytics/ultralyt
The YOLO command line interface (CLI) allows for simple single-line commands without the need for a Python environment.
The YOLO command line interface (CLI) allows for simple single-line commands without the need for a Python environment.
CLI requires no customization or Python code. You can simply run all tasks from the terminal with the `yolo` command. Check out the [CLI Guide](usage/cli.md) to learn more about using YOLOv8 from the command line.
CLI requires no customization or Python code. You can simply run all tasks from the terminal with the `yolo` command. Check out the [CLI Guide](usage/cli.md) to learn more about using YOLOv8 from the command line.
!!! example
!!! example
=== "Syntax"
=== "Syntax"
@ -93,7 +92,6 @@ CLI requires no customization or Python code. You can simply run all tasks from
yolo cfg
yolo cfg
```
```
!!! warning "Warning"
!!! warning "Warning"
Arguments must be passed as `arg=val` pairs, split by an equals `=` sign and delimited by spaces ` ` between pairs. Do not use `--` argument prefixes or commas `,` between arguments.
Arguments must be passed as `arg=val` pairs, split by an equals `=` sign and delimited by spaces ` ` between pairs. Do not use `--` argument prefixes or commas `,` between arguments.
description: Explore Ultralytics events, including 'request_with_credentials' and 'smart_request', to improve your project's performance and efficiency.
description: Detect 80+ object categories with bounding box coordinates and class probabilities using AutoShape in Ultralytics YOLO. Explore Detections now.
description: Learn how to register custom event-tracking and track predictions with Ultralytics YOLO via on_predict_start and register_tracker methods.
description: 'TrackState: A comprehensive guide to Ultralytics tracker''s BaseTrack for monitoring model performance. Improve your tracking capabilities now!'
description: '"Optimize tracking with Ultralytics BOTrack. Easily sort and track bots with BOTSORT. Streamline data collection for improved performance."'
description: '"Track Google Marketing Campaigns in GMC with Ultralytics Tracker. Learn to set up and use GMC for detailed analytics. Get started now."'
description: Learn how to match and fuse object detections for accurate target tracking using Ultralytics' YOLO merge_matches, iou_distance, and embedding_distance.
description: Learn how to use auto_annotate in Ultralytics YOLO to generate annotations automatically for your dataset. Simplify object detection workflows.
description: Learn about BaseDataset in Ultralytics YOLO, a flexible dataset class for object detection. Maximize your YOLO performance with custom datasets.
description: Convert COCO-91 to COCO-80 class, RLE to polygon, and merge multi-segment images with Ultralytics YOLO data converter. Improve your object detection.
description: 'Ultralytics YOLO Docs: Learn about stream loaders for image and tensor data, as well as autocasting techniques. Check out SourceTypes and more.'
description: Create custom YOLOv5 datasets with Ultralytics YOLODataset and SemanticDataset. Streamline your object detection and segmentation projects.
description: Learn about BaseTensor & Boxes in Ultralytics YOLO Engine. Check out Ultralytics Docs for quality tutorials and resources on object detection.
description: Dynamically adjusts input size to optimize GPU memory usage during training. Learn how to use check_train_batch_size with Ultralytics YOLO.
description: Improve your YOLOv5 model training with callbacks from ClearML. Learn about log debug samples, pre-training routines, validation and more.
description: Learn about YOLO callbacks using the Comet.ml platform, enhancing object detection training and testing with custom logging and visualizations.
description: Improve YOLOv5 model training with Ultralytics' on-train callbacks. Boost performance on-pretrain-routine-end, model-save, train/predict start.
description: '"Improve YOLO model performance with on_fit_epoch_end callback. Learn to integrate with Ray Tune for hyperparameter tuning. Ultralytics YOLO docs."'
description: Learn how to use Ultralytics YOLO's built-in callbacks `on_pretrain_routine_start` and `on_train_epoch_end` for improved training performance.
description: Learn about Bounding Boxes (Bboxes) and _ntuple in Ultralytics YOLO for object detection. Improve accuracy and speed with these powerful tools.
description: 'Discover the power of YOLO''s plotting functions: Colors, Labels and Images. Code examples to output targets and visualize features. Check it now.'
description: Improve your YOLO models with Ultralytics' TaskAlignedAssigner, select_highest_overlaps, and dist2bbox utilities. Streamline your workflow today.
description: Boost posture detection using PoseTrainer and train models using train() API. Learn PoseLoss for ultra-fast and accurate pose detection with Ultralytics YOLO.
description: '"Use SegmentationPredictor in YOLOv8 for efficient object detection and segmentation. Explore Ultralytics YOLO Docs for more information."'
description: Learn about SegmentationTrainer and Train in Ultralytics YOLO v8 for efficient object detection models. Improve your training with Ultralytics Docs.
Glenn Jocher
2 years ago
committed by
GitHub