The paper presents a dataset called FERET, which consists of a training set FA with a single face image for each of 1196 subjects, and 4 probe sets with pose, expression, illumination, and time-span variations.
BibTex:
Before browse our site, please accept our cookies policy
