• Sample Images. The U.S. National Institute of Standards & Technology (NIST) provided 3,368 digital facial test images. LINK: NIST Feret test facial images

  Batch Conversion of test facial images to Binary. All 3,368 NIST Feret test facial images were converted to Binary using proprietary software with permission.

  Biometric Face Recognition Testbed. Colorado State University (CSU) hosts a testbed for detailed performance testing of the leading biometric face recognition algorithms. LINK: Colorado State University: Evaluation of Face Recognition Algorithms

  Test Setup. Entire CSU testbed was installed on the BeOS 5.3 OS, a capable but relatively simple-to-operate OS now re-emerging as an Open Source OS named Haiku. LINK: CSU testbed installed on Haiku OS

  Control Run. A Control Run was performed using several of the leading biometric face recognition algorithms and the 3,368 NIST test facial images in Grayscale form. (To interpret the Result Charts, the higher the curve, the better the performance). The Results of the Grayscale Control Run are posted here. LINK: Control Run Results with Grayscale facial images

  Binary Run. The Control Run (above) was repeated, only this time Binary facial images were substituted in place of the Grayscale facial images. The results of the Binary Run are posted here. LINK: Test Run Results with Binary facial images