frec is a service that builds a database of recognizable faces upon time.
frec uses the AT&T Database of Faces for testing the service. It is free and available to anyone and credit goes to AT&T Laboratories Cambridge.
frec draws on the work of many bright individuals. These are some papers and or reference work that helped us along the way:
- Enhanced Local Texture Feature Sets for Face Recognition Under Difﬁcult Lighting Conditions - by Xiaoyang Tan and Bill Triggs
- Face Detection using LBP features - by Jo Chang-yeon
- Studies on Sensitivity of Face Recognition Performance to Eye Location Accuracy - Haoshu Wang - University of Notre Dame 2004
frec uses some math and these are articles that might help: