상관계수를 이용하여 인식률을 향상시킨 rank-level fusion 방법 = Rank-level Fusion Method That Improves Recognition Rate by Using Correlation Coefficient

다국어 초록 (Multilingual Abstract)

Currently, most biometrics system authenticates users by using single biometric information. This method has many problemssuch as noise problem, sensitivity to data, spoofing, a limitation of recognition rate. One method to solve this problems is to usemulti biometric information. The multi biometric authentication system performs information fusion for each biometric informationto generate new information, and then uses the new information to authenticate the user. Among information fusion methods, ascore-level fusion method is widely used. However, there is a problem that a normalization operation is required, and even ifdata is same, the recognition rate varies depending on the normalization method. A rank-level fusion method that does not requirenormalization is proposed. However, a existing rank-level fusion methods have lower recognition rate than score-level fusionmethods. To solve this problem, we propose a rank-level fusion method with higher recognition rate than a score-level fusionmethod using correlation coefficient. The experiment compares recognition rate of a existing rank-level fusion methods with therecognition rate of proposed method using iris information(CASIA V3) and face information(FERET V1). We also compare withscore-level fusion methods. As a result, the recognition rate improve from about 0.3% to 3.3%.

국문 초록 (Abstract)

현재 대부분의 생체인증 시스템은 단일 생체정보를 이용하여 사용자를 인증하고 있는데, 이러한 방식은 노이즈로 인한문제, 데이터에 대한 민감성 문제, 스푸핑, 인식률의 한계 등 많은 문제점들을 가지고 있다. 이를 해결하기 위한 방법 중하나로 다중 생체정보를 이용하는 방법이 제시되고 있다. 다중 생체인증 시스템은 각각의 생체정보에 대해서 informationfusion을 수행하여 새로운 정보를 생성한 뒤, 그 정보를 활용하여 사용자를 인증하는 방식이다. Information fusion방법들 중에서 score-level fusion 방법을 보편적으로 많이 사용한다. 하지만 정규화 작업이 필요하다는 문제점을 갖고있고, 데이터가 같아도 정규화 방법에 따라 인식률이 달라진다는 문제점을 갖고 있다. 이에 대한 대안으로 정규화 작업이필요 없는 rank-level fusion 방법이 제시되고 있다. 하지만 기존의 rank-level fusion 방법들은 score-level fusion방법보다 인식률이 낮다. 이러한 문제점을 해결하기 위해 상관계수를 이용하여 score-level fusion 방법보다 인식률이높은 rank-level fusion 방법을 제안한다. 실험은 홍채정보(CASIA V3)와 얼굴정보(FERET V1)를 이용하여 기존의존재하는 rank-level fusion 방법들의 인식률과 본 논문에서 제안하는 fusion 방법의 인식률을 비교하였다. 또한score-level fusion 방법들과도 인식률을 비교하였다. 그 결과로 인식률이 약 0.3%에서 3.3%까지 향상되었다.

참고문헌 (Reference)

1 M. Van Erp, "Variants of the borda count method for combining ranked classifier hypotheses" 443-452, 2000

2 D. Black, "The theory of committees and elections" Springer 1958

3 R. Sharma, "Rank level fusion in multibiometric systems" 1-4, 2015

4 A. Kumar, "Personal identification using multibiometrics rank-level fusion" 41 : 743-752, 2011

5 M. M. Monwar, "Multimodal biometric system using rank-level fusion approach" 39 : 867-878, 2009

6 S. Garcia-Salicetti, "Multimodal biometric score fusion: the mean rule vs. support vector classifiers" 1-4, 2005

7 O. Melnik, "Mixed group ranks: preference and confidence in classifier combination" 26 : 973-981, 2004

8 M.M. Monwar, "Intergrating monomodal biometric matchers through logistic regression rank aggregation approach" 1-7, 2008

9 A. Ross, "Information fusion in biometrics" 24 : 2115-2125, 2003

10 M. Jović, "Image retrieval based on similarity score fusion from feature similarity ranking lists" Sipinger 4223 : 461-470, 2006

11 J. Daugman, "How iris recognition works" 14 : 21-30, 2004

12 A. A. Ross, "Handbook of multibiometrics vol. 6" Springer Science & Business Media 2006

13 B. H. Guo, "Fusion analysis of soft biometrics for recognition at a distance" 1-8, 2018

14 B. Ammour, "Face-iris multi-modal biometric system using multi-resolution log-gabor filter with spectral regression kernel discriminant analysis" 7 : 482-489, 2018

15 M. Turk, "Eigenfaces for recognition" 3 : 77-86, 2002

16 Jung-Ho Ahn, "Effective rank-level fusion method to improve recognition rate" 2017

17 T. K. Ho, "Decision combination in multiple classifier systems" 16 : 66-75, 1994

18 N. Poh, "Database, protocols and tools for evaluating score-level fusion algorithms in biometric authentication" 39 : 222-223, 2006

19 J. K. Sing, "Confidence factor weighted Gaussian function induced parallel fuzzy rank-level fusion for inference and its application to face recognition" 47 : 60-71, 2019

20 H. Talebi, "Confidence based rank level fusion for multimodal biometric systems" Springer 9256 : 211-222, 2015

21 Ki-Young Moon, "Biometrics technology status and prospect" 98 (98): 38-47, 2005