Kvasir-SEG: Gastrointestinal Polyp Segmentation Dataset

Dataset Description

Kvasir-SEG is a gastrointestinal polyp segmentation dataset containing 1000 polyp images with corresponding segmentation masks from the Kvasir dataset.

• Task: Binary segmentation (polyp vs. background)
• Modality: Gastrointestinal endoscopy
• Format: PNG images (332x487 to 1920x1072 pixels) with binary masks
• Splits: Training, Validation, and Test sets

Dataset Structure

kvasir-seg/
├── train/
│   ├── images/  # Training images
│   └── masks/   # Training masks
├── validation/
│   ├── images/  # Validation images
│   └── masks/   # Validation masks
└── test/
    ├── images/  # Test images
    └── masks/   # Test masks

Citation

If you use this dataset, please cite:

@inproceedings{jha2020kvasir,
  title={Kvasir-SEG: A Segmented Polyp Dataset},
  author={Jha, Debesh and Smedsrud, Pia H and Riegler, Michael A and Halvorsen, P{a}l and de Lange, Thomas and Johansen, Dag and Johansen, H{a}vard D},
  booktitle={International Conference on Multimedia Modeling},
  pages={451--462},
  year={2020},
  organization={Springer}
}

Dataset Split: This train/validation/test split is created by the following study. Please find more details in:

@article{chang2024esfpnet,
  title={ESFPNet: Efficient Stage-Wise Feature Pyramid on Mix Transformer for Deep Learning-Based Cancer Analysis in Endoscopic Video},
  author={Chang, Qi and Ahmad, Danish and Toth, Jennifer and Bascom, Rebecca and Higgins, William E},
  journal={Journal of Imaging},
  volume={10},
  number={8},
  pages={191},
  year={2024},
  publisher={MDPI}
}

Usage

from datasets import load_dataset

# Load dataset
dataset = load_dataset("Angelou0516/kvasir-seg")

# Access splits
train_data = dataset['train']
val_data = dataset['validation']
test_data = dataset['test']

# Access a sample
sample = train_data[0]
image = sample['file_name']  # Image path
label = sample['label']  # Segmentation mask path

License

Please refer to the original Kvasir dataset license and citation requirements.

Links

• Original Dataset: https://datasets.simula.no/kvasir-seg/
• Paper: https://link.springer.com/chapter/10.1007/978-3-030-37734-2_37