COCO12-Formats Dataset

Introduction

The Ultralytics COCO12-Formats dataset is a specialized test dataset designed to validate image loading across all 12 supported image format extensions. It contains 12 images (6 for training, 6 for validation), each saved in a different format to ensure comprehensive testing of the image loading pipeline.

This dataset is invaluable for:

• Testing image format support: Verify that all supported formats load correctly
• CI/CD pipelines: Automated testing of format compatibility
• Debugging: Isolate format-specific issues in training pipelines
• Development: Validate new format additions or changes

Supported Formats

The dataset includes one image for each of the 12 supported format extensions defined in ultralytics/data/utils.py:

Dataset Structure

coco12-formats/
├── images/
│   ├── train/          # 6 images (avif, bmp, dng, heic, jpeg, jpg)
│   └── val/            # 6 images (jp2, mpo, png, tif, tiff, webp)
├── labels/
│   ├── train/          # Corresponding YOLO format labels
│   └── val/
└── coco12-formats.yaml # Dataset configuration

Dataset YAML

The COCO12-Formats dataset is configured using a YAML file that defines dataset paths and class names. You can review the official coco12-formats.yaml file in the Ultralytics GitHub repository.

# Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license

# COCO12-Formats dataset (12 images testing all supported image formats) by Ultralytics
# Documentation: https://docs.ultralytics.com/datasets/detect/coco12-formats/
# Example usage: yolo train data=coco12-formats.yaml
# parent
# ├── ultralytics
# └── datasets
#     └── coco12-formats ← downloads here (1 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: coco12-formats # dataset root dir
train: images/train # train images (relative to 'path') 6 images
val: images/val # val images (relative to 'path') 6 images
test: # test images (optional)

# Classes
names:
  0: person
  1: bicycle
  2: car
  3: motorcycle
  4: airplane
  5: bus
  6: train
  7: truck
  8: boat
  9: traffic light
  10: fire hydrant
  11: stop sign
  12: parking meter
  13: bench
  14: bird
  15: cat
  16: dog
  17: horse
  18: sheep
  19: cow
  20: elephant
  21: bear
  22: zebra
  23: giraffe
  24: backpack
  25: umbrella
  26: handbag
  27: tie
  28: suitcase
  29: frisbee
  30: skis
  31: snowboard
  32: sports ball
  33: kite
  34: baseball bat
  35: baseball glove
  36: skateboard
  37: surfboard
  38: tennis racket
  39: bottle
  40: wine glass
  41: cup
  42: fork
  43: knife
  44: spoon
  45: bowl
  46: banana
  47: apple
  48: sandwich
  49: orange
  50: broccoli
  51: carrot
  52: hot dog
  53: pizza
  54: donut
  55: cake
  56: chair
  57: couch
  58: potted plant
  59: bed
  60: dining table
  61: toilet
  62: tv
  63: laptop
  64: mouse
  65: remote
  66: keyboard
  67: cell phone
  68: microwave
  69: oven
  70: toaster
  71: sink
  72: refrigerator
  73: book
  74: clock
  75: vase
  76: scissors
  77: teddy bear
  78: hair drier
  79: toothbrush

# Download script/URL (optional)
download: https://github.com/ultralytics/assets/releases/download/v0.0.0/coco12-formats.zip

Dataset Generation

The dataset can be generated using the provided script that converts source images from COCO8 and COCO128 to all supported formats:

from ultralytics.data.scripts.generate_coco12_formats import generate_coco12_formats

# Generate the dataset
generate_coco12_formats()

Requirements

Some formats require additional dependencies:

pip install pillow pillow-heif pillow-avif-plugin

AVIF System Library (Optional)

For OpenCV to read AVIF files directly, libavif must be installed before building OpenCV:

brew install libavif

sudo apt install libavif-dev libavif-bin

git clone -b v1.2.1 https://github.com/AOMediaCodec/libavif.git
cd libavif
cmake -B build -DAVIF_CODEC_AOM=SYSTEM -DAVIF_BUILD_APPS=ON
cmake --build build --config Release --parallel
sudo cmake --install build

Usage

To train a YOLO model on the COCO12-Formats dataset, use the following examples:

from ultralytics import YOLO

# Load a pretrained YOLO model
model = YOLO("yolo26n.pt")

# Train on COCO12-Formats to test all image formats
results = model.train(data="coco12-formats.yaml", epochs=1, imgsz=640)

# Train YOLO on COCO12-Formats
yolo detect train data=coco12-formats.yaml model=yolo26n.pt epochs=1 imgsz=640

Format-Specific Notes

AVIF (AV1 Image File Format)

AVIF is a modern image format based on the AV1 video codec, offering excellent compression. Requires pillow-avif-plugin:

pip install pillow-avif-plugin

DNG (Digital Negative)

DNG is Adobe's open RAW format based on TIFF. For testing purposes, the dataset uses TIFF-based files with the .dng extension.

JP2 (JPEG 2000)

JPEG 2000 is a wavelet-based image compression standard offering better compression and quality than traditional JPEG. Commonly used in medical imaging (DICOM), geospatial applications, and digital cinema. Natively supported by both OpenCV and Pillow.

MPO (Multi-Picture Object)

MPO files are used for stereoscopic (3D) images. The dataset stores standard JPEG data with the .mpo extension for format testing.

HEIC (High Efficiency Image Coding)

HEIC requires the pillow-heif package for proper encoding:

pip install pillow-heif

Use Cases

CI/CD Testing

from ultralytics import YOLO


def test_all_image_formats():
    """Test that all image formats load correctly."""
    model = YOLO("yolo26n.pt")
    results = model.train(data="coco12-formats.yaml", epochs=1, imgsz=64)
    assert results is not None

Format Validation

from pathlib import Path

from ultralytics.data.utils import IMG_FORMATS

# Verify all formats are represented
dataset_dir = Path("datasets/coco12-formats/images")
found_formats = {f.suffix[1:].lower() for f in dataset_dir.rglob("*.*")}
assert found_formats == IMG_FORMATS, f"Missing formats: {IMG_FORMATS - found_formats}"

Citations and Acknowledgments

If you use the COCO dataset in your research, please cite:

@misc{lin2015microsoft,
      title={Microsoft COCO: Common Objects in Context},
      author={Tsung-Yi Lin and Michael Maire and Serge Belongie and Lubomir Bourdev and Ross Girshick and James Hays and Pietro Perona and Deva Ramanan and C. Lawrence Zitnick and Piotr Doll{\'a}r},
      year={2015},
      eprint={1405.0312},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

FAQ

What Is the COCO12-Formats Dataset Used For?

The COCO12-Formats dataset is designed for testing image format compatibility in Ultralytics YOLO training pipelines. It ensures all 12 supported image formats (AVIF, BMP, DNG, HEIC, JP2, JPEG, JPG, MPO, PNG, TIF, TIFF, WebP) load and process correctly.

Why Test Multiple Image Formats?

Different image formats have unique characteristics (compression, bit depth, color spaces). Testing all formats ensures:

• Robust image loading code
• Compatibility across diverse datasets
• Early detection of format-specific bugs

Which Formats Require Special Dependencies?

• AVIF: Requires pillow-avif-plugin
• HEIC: Requires pillow-heif

Can I Add New Format Tests?

Yes! Modify the generate_coco12_formats.py script to include additional formats. Ensure you also update IMG_FORMATS in ultralytics/data/utils.py.

📅 Created 0 days ago ✏️ Updated 0 days ago

Comments