Ultralytics COCO8-pose Dataset

Ultralytics COCO8-pose is a small, but versatile pose detection dataset composed of the first 8 images of the COCO train 2017 set, 4 for training and 4 for validation. This dataset is ideal for testing and debugging object detection models, or for experimenting with new detection approaches. With 8 images, it is small enough to be easily manageable, yet diverse enough to test training pipelines for errors and act as a sanity check before training… See the full description on the dataset page: https://huggingface.co/datasets/Ultralytics/COCO8-pose.

Authors marked with an asterisk (*) have contributed equally to this publication.

We annotated a dataset for the detection of drainage outlets and ventilations on flat rooftops. The underlying high-resolution aerial images are orthophotos with a ground sampling distance of 7.5 cm, provided by the Office for Land Management and Geoinformation of the City of Bonn, Germany. The dataset was created through manual annotation using the Computer Vision Annotation Tool (CVAT) and comprises 740 image pairs. Each pair consists of a rooftop image and a corresponding annotated mask indicating the drainage outlets and ventilations. Since rooftops vary in size, we aimed to create image pairs that capture a single rooftop per image without overlaps or cutoffs. Consequently, the dimensions of each image pair differ. The dataset is split randomly into 80% for training, 10% for validation, and 10% for testing.

We provide the dataset in the Common Objects in Context (COCO) format for object detection tasks. In addition to the COCO-formatted dataset, we provide the dataset in its original, pairwise, format to support various machine learning tasks, such as semantic segmentation and panoptic segmentation, as well as to accommodate different data-loading requirements for diverse deep learning models.

If your object detection approach requires the 'category_id' to start from 0 instead of 1, please refer to the following guide: https://github.com/obss/sahi/discussions/336
For conversion to a completely different dataset format, such as YOLO, please see the repository: https://github.com/ultralytics/JSON2YOLO

This dataset contains labelled underwater pictures taken at the OBSEA underwater observatory. The pictures have already been splited into train, validation and test folders for training a YOLO AI model. This is a substet of this dataset, with the data arranged to simplify the training process.

Train a YOLO model with this dataset:

1. Install dependencies with the command "pip3 install ultralytics"
2. Download and unzip the file
3. Edit the data.yaml file to match the "train", "valid" and "test" paths in your filesystem
4. Train the model with the command: "yolo task=detect mode=train model=yolov8n.pt data=data.yaml"

Done! In order to optimize the results it is encouraged to tune the hyperparameters to fit your application.

Bangla LPDB - A (YOLO Ready)

This dataset is a reformatted and enhanced version of the Bangla LPDB - A dataset, originally published by Ataher Sams and Homaira Huda Shomee. It has been meticulously prepared to be plug-and-play for YOLO (You Only Look Once) object detection models, making it incredibly easy for researchers and developers to use for license plate detection tasks in Bangladeshi vehicles.

Original Dataset Acknowledgment

This dataset is built upon

Ataher Sams, & Homaira Huda Shomee. (2021). Bangla LPDB - A (Version v1) [Data set]. International Conference on Digital Image Computing: Techniques and Applications (IEEE DICTA), Gold Coast, Queensland Australia. Zenodo. https://doi.org/10.5281/zenodo.4718238

We extend our sincerest gratitude to them for creating such a comprehensive and vital resource for the research community.

Key Enhancements

While the original Bangla LPDB - A dataset is an excellent collection, this version provides significant improvements for immediate use with YOLO models:

• YOLO Dataset Structure: The dataset has been meticulously restructured to follow the standard YOLO format, including all necessary annotation files.
• dataset.yaml Included: A dataset.yaml file is provided for seamless integration with popular deep learning frameworks like Ultralytics YOLO.
• Class Names: Separate files for both Bengali and English class names are included, offering flexibility for various applications.
• Pre-split Data: The dataset is pre-split into optimal ratios for training, validation, and testing:
  • Training Set: 70%
  • Validation Set: 20%
  • Testing Set: 10%

Preview

Vehicle to License Plate https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F24265111%2F3976ff0d5a259dbd70dc017964bf7d47%2Fvehicle-to-license-plate.png?generation=1753064814234359&alt=media" alt="">

License Plate to Text https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F24265111%2F8f5461031162387a7aba8ed7f31c6eda%2Flicense-plate-to-text.png?generation=1753064874309479&alt=media" alt="">

The Original Paper

H. H. Shomee and A. Sams, "License Plate Detection and Recognition System for All Types of Bangladeshi Vehicles Using Multi-step Deep Learning Model," 2021 Digital Image Computing: Techniques and Applications (DICTA), Gold Coast, Australia, 2021, pp. 01-07, https://doi.org/10.1109/DICTA52665.2021.9647284.

Dataset Citation

Users of this dataset are required to cite the original research paper, which introduces the Bangla LPDB - A dataset and its applications. Please use the following citation:

@INPROCEEDINGS{9647284,
 author={Shomee, H. H. and Sams, A.},
 booktitle={2021 Digital Image Computing: Techniques and Applications (DICTA)},
 title={License Plate Detection and Recognition System for All Types of Bangladeshi Vehicles Using Multi-step Deep Learning Model},
 year={2021},
 pages={01-07},
 doi={10.1109/DICTA52665.2021.9647284}
}

Modified by

Ashikur Rahman Shad

ashikur-rahman-shad (Ashikur Rahman Shad) · GitHub

Eagle Dataset - YOLOv11 OBB for Vehicle Detection

High-resolution aerial imagery with 16,000+ oriented bounding boxes for vehicle detection, pre-formatted for Ultralytics YOLOv11.

Context

This dataset is a ready-to-use version of the original Eagle Dataset from the German Aerospace Center (DLR). The original dataset was created to benchmark object detection models on challenging aerial imagery, featuring vehicles at various orientations.

This version has been converted to the YOLOv11-OBB (Oriented Bounding Box) format. The conversion makes the dataset directly compatible with modern deep learning frameworks like Ultralytics YOLO, allowing researchers and developers to train state-of-the-art object detectors with minimal setup.

The dataset is ideal for tasks requiring precise localization of rotated objects, such as vehicle detection in parking lots, traffic monitoring, and urban planning from aerial viewpoints.

Content

The dataset is split into training, validation, and test sets, following a standard structure for computer vision tasks.

Dataset Split & Counts:

• Training Set: 159 images and labels
• Validation Set: 53 images and labels
• Test Set: 106 images and labels

Directory Structure:

EagleDatasetYOLO/
├── train/
│  ├── images/   # 159 images
│  └── labels/   # 159 .txt obb labels
├── val/
│  ├── images/   # 53 images
│  └── labels/   # 53 .txt obb labels
├── test/
│  ├── images/   # 106 images
│  └── labels/   # 106 .txt obb labels
├── data.yaml
└── license.md

Annotation Format (YOLOv11-OBB):

Each .txt label file contains one object per line. The format for each object is: <class_id> <x_center> <y_center> <width> <height> <angle>

• <class_id>: The class index (in this case, 0 for 'vehicle').
• <x_center> <y_center>: The normalized center coordinates of the bounding box.
• <width> <height>: The normalized width and height of the bounding box.
• <angle>: The rotation angle of the box in radians, from -π/2 to π/2.

data.yaml Configuration:

A data.yaml file is included for easy integration with the Ultralytics framework.

path: ../EagleDatasetYOLO
train: train/images
val: val/images
test: test/images

nc: 1
names: ['vehicle']

Acknowledgements and License

This dataset is a conversion of the original work by the German Aerospace Center (DLR). The conversion to YOLOv11-OBB format was performed by Mridankan Mandal.

The dataset is released under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International license (CC BY-NC-SA 4.0).

If you use this dataset in your research, please cite the original creators and acknowledge the conversion work.

The most vulnerable group of traffic participants are pedestrians using mobility aids. While there has been significant progress in the robustness and reliability of camera based general pedestrian detection systems, pedestrians reliant on mobility aids are highly underrepresented in common datasets for object detection and classification.

To bridge this gap and enable research towards robust and reliable detection systems which may be employed in traffic monitoring, scheduling, and planning, we present this dataset of a pedestrian crossing scenario taken from an elevated traffic monitoring perspective together with ground truth annotations (Yolo format [1]). Classes present in the dataset are pedestrian (without mobility aids), as well as pedestrians using wheelchairs, rollators/wheeled walkers, crutches, and walking canes. The dataset comes with official training, validation, and test splits.

An in-depth description of the dataset can be found in [2]. If you make use of this dataset in your work, research or publication, please cite this work as:

@inproceedings{mohr2023mau,
author = {Mohr, Ludwig and Kirillova, Nadezda and Possegger, Horst and Bischof, Horst},
title = {{A Comprehensive Crossroad Camera Dataset of Mobility Aid Users}},
booktitle = {Proceedings of the 34th British Machine Vision Conference ({BMVC}2023)},
year = {2023}
}

Archive mobility.zip contains the full detection dataset in Yolo format with images, ground truth labels and meta data, archive mobility_class_hierarchy.zip contains labels and meta files (Yolo format) for training with class hierarchy using e.g. the modified version of Yolo v5/v8 available under [3].
To use this dataset with Yolo, you will need to download and extract the zip archive and change the path entry in dataset.yaml to the directory where you extracted the archive to.

[1] https://github.com/ultralytics/ultralytics
[2] coming soon
[3] coming soon

This database contains 4976 planetary images of boulder fields located on Earth, Mars and Moon. The data was collected during the BOULDERING Marie Skłodowska-Curie Global fellowship between October 2021 and 2024. The data was already splitted into train, validation and test datasets, but feel free to re-organize the labels at your convenience.

For each image, all of the boulder outlines within the image were carefully mapped in QGIS. More information about the labelling procedure can be found in the following manuscript (https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2023JE008013). This dataset differs from the previous dataset included along with the manuscript https://zenodo.org/records/8171052, as it contains more mapped images, especially of boulder populations around young impact structures on the Moon (cold spots). In addition, the boulder outlines were also pre-processed so that it can be ingested directly in YOLOv8.

A description of what is what is given in the README.txt file (in addition in how to load the custom datasets in Detectron2 and YOLO). Most of the other files are mostly self-explanatory. Please see previous dataset or manuscript for more information. If you want to have more information about specific lunar and martian planetary images, the IDs of the images are still available in the name of the file. Use this ID to find more information (e.g., M121118602_00875_image.png, ID M121118602 ca be used on https://pilot.wr.usgs.gov/). I will also upload the raw data from which this pre-processed dataset was generated (see https://zenodo.org/records/14250970).

Thanks to this database, you can easily train a Detectron2 Mask R-CNN or YOLO instance segmentation models to automatically detect boulders.

How to cite:

Please refer to the "how to cite" section of the readme file of https://github.com/astroNils/YOLOv8-BeyondEarth.

Structure:

. └── boulder2024/ ├── jupyter-notebooks/ │ └── REGISTERING_BOULDER_DATASET_IN_DETECTRON2.ipynb ├── test/ │ └── images/ │ ├── _image.png │ ├── ... │ └── labels/ │ ├── _image.txt │ ├── ... ├── train/ │ └── images/ │ ├── _image.png │ ├── ... │ └── labels/ │ ├── _image.txt │ ├── ... ├── validation/ │ └── images/ │ ├── _image.png │ ├── ... │ └── labels/ │ ├── _image.txt │ ├── ... ├── detectron2_inst_seg_boulder_dataset.json ├── README.txt ├── yolo_inst_seg_boulder_dataset.yaml

detectron2_inst_seg_boulder_dataset.json

is a json file containing the masks as expected by Detectron2 (see https://detectron2.readthedocs.io/en/latest/tutorials/datasets.html for more information on the format). In order to use this custom dataset, you need to register the dataset before using it in the training. There is an example how to do that in the jupyter-notebooks folder. You need to have detectron2, and all of its depedencies installed.

yolo_inst_seg_boulder_dataset.yaml

can be used as it is, however you need to update the paths in the .yaml file, to the test, train and validation folders. More information about the YOLO format can be found here (https://docs.ultralytics.com/datasets/segment/).

iSAID-YOLO11-Seg: A Cleaned YOLO11 Segmentation Conversion of the iSAID Aerial Instance Segmentation Dataset

Overview A large-scale aerial imagery dataset reformatted from the original iSAID instance segmentation benchmark into Ultralytics’ YOLO11 segmentation format. Polygon masks have been normalized and cleaned for seamless training, validation, and inference with YOLO11-seg models .

Dataset Composition

• Total Images:

  • Train: 28,029
  • Validation: 9,512
  • Test (unlabeled): 19,377

• Total Annotated Instances: 655,451 across 15 object categories (plus one “unlabeled” class for background).

• Image Resolution: Original 6000×6000 tiles split into 800×800 crops.

• Annotation Files: YOLO11 .txt segmentation labels (normalized coordinates).

Directory Structure

iSAID-YOLO11-Seg/
├── images/
│  ├── train/    # 28,029 images
│  ├── val/     # 9,512 images
│  └── test/    # 19,377 images
├── labels/
│  ├── train/    # 28,029 .txt polygon labels
│  └── val/     # 9,512 .txt polygon labels
├── data.yaml    # Dataset configuration
├── ReadMe.md    # Dataset description and usage
└── license.md    # License terms :contentReference[oaicite:2]{index=2}

Annotation Format

Each label file contains one polygon instance per line:

Context

One of my passions is playing board games with my friends. However one of them lives abroad and so we like to stream the game when playing with him. However instead of just having a normal stream, I wanted to show some additional information about the monsters that are on the game board. This originated in a fun project to train CNNs in order to detect these monsters.

Content

To have enough training data, I made a little project in UE4 to generate these training images. For each image there is a mask for every monster that appears in it. The dataset also includes annotations for the train images in the COCO format (annotations.json) and labes for the bounding box in Darknet format in the folder labels.

There is a training and validation subset for the images, labels and masks folders. The structure is as follows: for the first training image containing an earth_demon and harrower_infester:

• The image is stored at images/train/image_1.png
• The label-file is stored at labels/train/label_1.png. This file contains two lines. One line for each monster. A line is constructed as follows: class_id center_x center_y width height. Note that the position and dimensions are relative to the image width and height.
• There are two mask images located at masks/train. One is named image_1_mask_0_harrower_infester.png and the other image_1_mask_1_earth_demon.png.

The code for generating this dataset and training a MaskRCNN and YoloV5 model can be found at https://github.com/ericdepotter/Gloomhaven-Monster-Recognizer.

Acknowledgements

I took pictures for the images of the monsters myself. The images of the game tiles I obtained from this collection of Gloomhaven assets.

Inspiration

This is a classic object detection or object segmentation problem.