Including annotations for each images in both forensic and caricature database.
We build our attribute dataset by annotating the caricature dataset and forensic dataset.
Caricature dataset we used is provided by Brendan F. Klare et al. from Michigan State University.
Forensic sketch dataset we used is obtained from IIITD forensic sketch database.
We annotate 77 attributes, 67 of them are also provided by Brendan F. Klare et al. from Michigan State University and we provide 10 extra attributes: cheeks, wrinkles on one's forehead, glasses, ties, teeth, race(black,white and yellow), blond hair and gender(male or female).