Description

The MSU PrintsGAN Dataset, was created to synthesize highly realistic, rolled fingerprint impressions with multiple impressions per generated identitiy. Currently, it consists of 35,000 unique, syntheticly generated fingerprint identities with 15 impressions each, for a total of 525,000 images. Fig. 1 presents some example PrintsGAN fingerprint images. The database was produced at Michigan State University’s Pattern Recognition and Image Processing (PRIP) Lab, in East Lansing, MI, US. This dataset may be augmented later with additional synthetic fingerprint identities and more impressions per identity.

Figure 1: Example images from the MSU PrintsGAN Dataset. The top row shows genuine pairs of the same finger, whereas the bottom row shows imposter pairs of different fingers. The minutiae correspondence and the match similarity scores using a commercial fingerprint matcher are shown for each pair.

Details

Download Instructions

To download the MSU PrintsGAN dataset, please first print out, fill and sign the Agreement and send it to: Steven Grosz (groszste@msu.edu). You will receive a download link upon approval of your usage of the database.

Acknowledgements

If you use this database, please cite the following publication: 


@article{engelsma2022printsgan,
  title={PrintsGAN: Synthetic Fingerprint Generator},
  author={Engelsma, Joshua J and Grosz, Steven A and Jain, Anil K},
  journal={arXiv preprint arXiv:2201.03674},
  year={2022}
}