BharathK333/DOOMGAN-Ocular-Morphs

DOOMGAN: Ocular Morph Dataset

This repository contains the official public dataset for the paper: "DOOMGAN: High-Fidelity Dynamic Identity Obfuscation Ocular Generative Morphing" funded by the NSF award no. 2345561.

The dataset consists of 10,000 high-fidelity morphed ocular images generated by the DOOMGAN model. These images are intended to facilitate research and development of Morph Attack Detection (MAD) systems for visible-spectrum ocular biometrics.

• Paper: IJCB 2025 DOOMGAN
• Model & Code: BharathK333/DOOMGAN
• Interactive Demo: DOOMGAN Space

Dataset Details

This dataset contains 10,000 morphed images with the following composition:

• 5,000 left-centric ocular morphs: Filenames contain 'l' and '_morph'.
• 5,000 right-centric ocular morphs: Filenames contain 'r' and '_morph'.

The images were generated by combining ocular images from two different identities, creating a synthetic biometric sample that can potentially be matched to either contributing individual.

Naming Convention

The filenames follow the pattern: 'id1_id2_morph.png'.

• 'id1', 'id2': Subject identifiers of the contributing identities.
• 'eye': Indicates 'l' for left eye or 'r' for right eye.
• 'session': Session number from the source dataset.

Example: '1141_l_1_1142_l_1_morph.png' is a morph of the left eye from subject '1141' and subject '1142'.

Using with the 'datasets' Library

You can easily load this dataset using the Hugging Face 'datasets' library. The library will automatically create splits based on the subfolder names.

from datasets import load_dataset

# Load the dataset
# This will create two splits: 'left_eye_morphs' and 'right_eye_morphs'
dataset = load_dataset("BharathK333/DOOMGAN-Ocular-Morphs")

print(dataset)

Intended Use

This dataset is created to support the academic and research communities. The primary intended use is to train and evaluate Morph Attack Detection (MAD) algorithms to defend biometric systems against this vulnerability.

Note: Please contact the authors for the complete dataset

Citation

If you use this dataset or our model in your work, please cite our paper:

@article{krishnamurthy2025doomgan,
  title={DOOMGAN: High-Fidelity Dynamic Identity Obfuscation Ocular Generative Morphing},
  author={Krishnamurthy, Bharath and Rattani, Ajita},
  journal={arXiv preprint arXiv:2507.17158},
  year={2025}
}

Acknowledgements

This work is supported in part by the National Science Foundation (NSF), United States award no. 2345561.