Facial Feature Verification System based on SVM Classifier


The KIPS Transactions:PartB , Vol. 11, No. 6, pp. 675-682, Oct. 2004
10.3745/KIPSTB.2004.11.6.675,   PDF Download:

Abstract

With the five-day workweek system in bank and the increased usage of ATM(Automatic Teller Machine), it is required that the financial crime using stolen credit card should be prevented. Though a CCTV camera is usually installed in near ATM, an intelligent criminal can cheat it disguising himself with sunglass or mask. In this paper, we propose facial feature verification system which can detect whether the user's face can be identified or not, using image processing algorithm and SVM(Support Vector Machine). Experimental results show that FAR(Error Rate for accepting a disguised man as a non-disguised one) is 1% and FRR(Error Rate for rejecting a normal/non-disguised man as a disguised one) is 2% for training data. In addition, it shows the FAR of 2.5% and the FRR of 1.43% for test data.


Statistics
Show / Hide Statistics

Statistics (Cumulative Counts from September 1st, 2017)
Multiple requests among the same browser session are counted as one view.
If you mouse over a chart, the values of data points will be shown.


Cite this article
[IEEE Style]
K. R. Park, J. H. Kim, S. Y. Lee, "Facial Feature Verification System based on SVM Classifier," The KIPS Transactions:PartB , vol. 11, no. 6, pp. 675-682, 2004. DOI: 10.3745/KIPSTB.2004.11.6.675.

[ACM Style]
Kang Ryoung Park, Jai Hie Kim, and Soo Youn Lee. 2004. Facial Feature Verification System based on SVM Classifier. The KIPS Transactions:PartB , 11, 6, (2004), 675-682. DOI: 10.3745/KIPSTB.2004.11.6.675.