Face Databases for 2D and 3D Facial Recognition: A Survey

Abstract

This paper presents a detailed case study of different face databases frequently used for face recognition and object recognition. The simple facedatabasedeveloped at AT & T labcontains 40 subjects each with 10 images. The database was collected by AT & T lab and it has the subjects of different pose and facial expressions. The yale database contains 15 subjects each with 11 images.The facial recognition technology database contain more than 4,000 both gray scale and colour images. Different 2D face recognition algorithms usedthis database fortheirperformance comparison.The cmu pie and multi-pie databases were developed for testing the performance ofdifferent videoand stillface recognition algorithms.The photometric stereo (PS) is a method to rapidly estimate the three-dimensionalgeometry of a face (or any other Lambertian-like object) using several images with an identical viewpoint but varied illumination directions.This paper work help to select appropriate face databases for both video face and still face recognition and also it gives an idea about selection of databases for 2D and 3D face recognition algorithms.