FRGC stands for Face Recognition Grand Challenge. The dataset consists of 50,000 images, divided into training set and validation set. The training set is used to train the algorithm, and the validation set is used to evaluate the performance of the method in a laboratory environment. The validation set contains data from 4,003 subject sessions.

A Subject session is a collection of all images of a person taken each time their biometric data is collected. It consists of four controlled still images, two uncontrolled still images, and one 3D image. The controlled images are taken in a studio and are full frontal images of two facial expressions (smiling and neutral) under two lighting conditions. The 3D images consist of range and texture images. The 3D images are acquired with a Minolta Vivid 900/910 series sensor.