DOTA128 Dataset
Introduction
Ultralytics DOTA128 is a small but versatile oriented object detection dataset composed of 128 images from the DOTAv1 set, 128 for training and validation. This dataset is ideal for testing and debugging oriented bounding box (OBB) models, or for experimenting with new detection approaches. With 128 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 larger datasets.
Dataset Structure
- Images: 128 aerial tiles (all in train folder, used for both train and val) sourced from DOTAv1.
- Classes: Inherits the 15 DOTAv1 categories such as plane, ship, and large vehicle.
- Labels: YOLO-format oriented bounding boxes saved as
.txtfiles beside each image.
This dataset is intended for use with Ultralytics Platform and YOLO26.
Dataset YAML
A YAML (Yet Another Markup Language) file is used to define the dataset configuration. It contains information about the dataset's paths, classes, and other relevant information. In the case of the DOTA128 dataset, the dota128.yaml file is maintained at https://github.com/ultralytics/ultralytics/blob/main/ultralytics/cfg/datasets/dota128.yaml.
ultralytics/cfg/datasets/dota128.yaml
# Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
# DOTA128 dataset (128 images from the DOTAv1 split) by Ultralytics
# Documentation: https://docs.ultralytics.com/datasets/obb/dota128/
# Example usage: yolo train model=yolo26n-obb.pt data=dota128.yaml
# parent
# ├── ultralytics
# └── datasets
# └── dota128 ← downloads here (34 MB)
# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]
path: dota128 # dataset root dir
train: images/train # train images (relative to 'path') 128 images
val: images/train # val images (relative to 'path') 128 images
# Classes for DOTA 1.0
names:
0: plane
1: ship
2: storage tank
3: baseball diamond
4: tennis court
5: basketball court
6: ground track field
7: harbor
8: bridge
9: large vehicle
10: small vehicle
11: helicopter
12: roundabout
13: soccer ball field
14: swimming pool
# Download script/URL (optional)
download: https://github.com/ultralytics/assets/releases/download/v0.0.0/dota128.zip
Usage
To train a YOLO26n-obb model on the DOTA128 dataset for 100 epochs with an image size of 640, you can use the following code snippets. For a comprehensive list of available arguments, refer to the model Training page.
Train Example
from ultralytics import YOLO
# Load a model
model = YOLO("yolo26n-obb.pt") # load a pretrained model (recommended for training)
# Train the model
results = model.train(data="dota128.yaml", epochs=100, imgsz=640)
# Start training from a pretrained *.pt model
yolo obb train data=dota128.yaml model=yolo26n-obb.pt epochs=100 imgsz=640
Sample Images and Annotations
Here are some examples of images from the DOTA128 dataset, along with their corresponding annotations:
- Mosaiced Image: This image demonstrates a training batch composed of mosaiced dataset images. Mosaicing is a technique used during training that combines multiple images into a single image to increase the variety of objects and scenes within each training batch. This helps improve the model's ability to generalize to different object sizes, aspect ratios, and contexts.
The example showcases the variety and complexity of the images in the DOTA128 dataset and the benefits of using mosaicing during the training process.
Citations and Acknowledgments
If you use the DOTA dataset in your research or development work, please cite the following paper:
@article{9560031,
author={Ding, Jian and Xue, Nan and Xia, Gui-Song and Bai, Xiang and Yang, Wen and Yang, Michael and Belongie, Serge and Luo, Jiebo and Datcu, Mihai and Pelillo, Marcello and Zhang, Liangpei},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
title={Object Detection in Aerial Images: A Large-Scale Benchmark and Challenges},
year={2021},
volume={},
number={},
pages={1-1},
doi={10.1109/TPAMI.2021.3117983}
}
A special note of gratitude to the team behind the DOTA datasets for their commendable effort in curating this dataset. For an exhaustive understanding of the dataset and its nuances, please visit the official DOTA website.
FAQ
What is the DOTA128 dataset and how can it be used?
The DOTA128 dataset is a versatile oriented object detection dataset made up of 128 images from the DOTAv1 set, all stored in the train folder. Both training and validation use the same set of images, making it ideal for quick testing and debugging workflows. It's ideal for testing and debugging OBB models like Ultralytics YOLO26. Due to its manageable size and diversity, it helps in identifying pipeline errors and running sanity checks before deploying larger datasets. Learn more about OBB detection with Ultralytics YOLO26.
How do I train a YOLO26 model using the DOTA128 dataset?
To train a YOLO26n-obb model on the DOTA128 dataset for 100 epochs with an image size of 640, you can use the following code snippets. For comprehensive argument options, refer to the model Training page.
Train Example
from ultralytics import YOLO
# Load a model
model = YOLO("yolo26n-obb.pt") # load a pretrained model (recommended for training)
# Train the model
results = model.train(data="dota128.yaml", epochs=100, imgsz=640)
# Start training from a pretrained *.pt model
yolo obb train data=dota128.yaml model=yolo26n-obb.pt epochs=100 imgsz=640
What are the key features of the DOTA dataset and where can I access the YAML file?
The DOTA dataset is known for its large-scale benchmark and the challenges it presents for object detection in aerial images. The DOTA128 subset provides more diversity than DOTA8 while remaining manageable for initial tests. You can access the dota128.yaml file, which contains paths, classes, and configuration details, at this GitHub link.
How does DOTA128 compare to other DOTA dataset variants?
DOTA128 (128 images) sits between DOTA8 (8 images) and the full DOTA-v1 dataset (1,869 images) in terms of size:
- DOTA8: Contains just 8 images (4 train, 4 val) - ideal for quick tests and debugging
- DOTA128: Contains 128 images (all in train folder, used for both train and val) - balanced between size and diversity
- Full DOTA-v1: Contains 1,869 images - comprehensive but resource-intensive
DOTA128 provides a good middle ground, offering more diversity than DOTA8 while remaining much more manageable than the full DOTA dataset for experimentation and initial model development.
How does mosaicing enhance model training with the DOTA128 dataset?
Mosaicing combines multiple images into one during training, increasing the variety of objects and contexts within each batch. This improves a model's ability to generalize to different object sizes, aspect ratios, and scenes. This technique can be visually demonstrated through a training batch composed of mosaiced DOTA128 dataset images, helping in robust model development. Explore more about mosaicing and training techniques on our Training page.
Why should I use Ultralytics YOLO26 for oriented object detection tasks?
Ultralytics YOLO26 provides state-of-the-art real-time object detection capabilities, including features like oriented bounding boxes (OBB), instance segmentation, and a highly versatile training pipeline. It's suitable for various applications and offers pretrained models for efficient fine-tuning. Explore further about the advantages and usage in the Ultralytics YOLO26 documentation.
