|
V1 N1 Paper 2
|
Annals of the MS in Computer Science and Information Systems at
UNC Wilmington
|
Spring 2007
|
Neurocognitive Inspired Hierarchical Face Recognition System
Ryan Wilkins
Committee
Abstract
The race-specific face recognition system introduced in this paper increases identification rates relative to the Eigenface baseline established by Mathew Turk. The race-specific system performs processes similar to those that humans do when identifying a face. The neurocognitive phenomenon identified as own-race bias is the basis of this system which first classifies the face by race and then identifies it against a stored database. Evaluation phase of the race-specific system compares the baseline recognition system against the race-specific by examining the methods which improve classification rates and identification rates. The comparison will evaluate the total system using a standard rank N identification rates. The MORPH database is used as the face corpus where particular images were sampled based on criteria. The MORPH database was developed at UNCW for analysis of the effects of age-progression.
download
(pdf)
Recommended Citation:
Wilkins, R., Ricanek, K, Vetter, R., Yaylacicegi, U. (2007) Neurocognitive Inspired Hierarchical Face Recognition System. Annals of the Master of Science in Computer Science and Information Systems at UNC Wilmington, 1(1) paper 2. http://csbapp.uncw.edu/data/mscsis/full.aspx.
|
V1 N1 Paper 2
|
Annals of the MS in Computer Science and Information Systems at
UNC Wilmington
|
Spring 2007
|
