Database Protocol for face photo-sketch synthesis and recognition algorithms

To enable the easy comparison of new face photo-sketch synthesis and recognition algorithms with other previously published methods, the protocol used to train and test methods in [1] and [2] is provided in the archive which can be downloaded by clicking here.

As detailed in the ‘readme’ file, you will find (i) the image file names of the face photo images used to populate the extended gallery (used in both [1] and [2]), (ii) the image file names of the face photos and sketches used to train and test algorithms when using viewed hand-drawn composite (VHDC) sketches (used in [2]), and (iii) the image file names of the face photos and sketches used to train and test algorithms when using viewed software-generated sketches in the UoM-SGFS database (used in [1]).

You will also find MATLAB ‘.mat’ files containing image file names of the photo-sketch pairs used and gender and ethnicity demographic information. While demographic information was NOT used to obtain the results in [1], [2], they may prove useful for researchers working in the field.

 

[1] C. Galea and R. A. Farrugia, “Matching Software-Generated Sketches to Face Photos with a Very Deep CNN, Morphed Faces, and Transfer Learning,” IEEE Transactions on Information Forensics and Security, vol. 13, no. 6, pp. 1421-1431, Jun. 2018.

BibTeX:
@article{Gal18,
author={Galea, C. and Farrugia, R. A.},
title={Matching Software-Generated Sketches to Face Photos with a Very Deep CNN, Morphed Faces, and Transfer Learning},
journal={IEEE Transactions on Information Forensics and Security},
year={2018},
volume={13},
number={6},
pages={1421-1431},
month={June},
}

 
[2] C. Galea and R. A. Farrugia, “Forensic face photo-sketch recognition using a deep learning-based architecture,” IEEE Signal Processing Letters, vol. 24, no. 11, pp. 1586-1590, Nov. 2017.

BibTeX:
@article{Gal17,
author={Galea, C. and Farrugia, R. A.},
title={Forensic face photo-sketch recognition using a deep learning-based architecture},
booktitle={IEEE Signal Processing Letters},
year={2017},
volume={24},
number={11},
pages={1586-1590},
month={November},
}