Facing the Challenge
- Wednesday, 20 June 2007
The recently released results of the Face Recognition Grand Challenge (FRGC), sponsored by the National Institute of Standards and Technology (NIST), indicate that the best face-recognition algorithms perform more accurately than most humans. The FRGC was open to researchers and developers in industry, academia, and research institutions, including Carnegie Mellon, Toshiba, Samsung, Google, and Cambridge University.
The FRGC was developed to advance the state of the art in face recognition technology, which has experienced renewed interest by the U.S. government and industry as a security measure. The challenge involved progressively difficult problems constructed around a data set of high-resolution still images, 3D images, and multiple images of a person. A defined set of experiments was performed to determine the performance of each algorithm.