Combining voiceprint and face biometrics for speaker identification using SDWS
暂无分享,去创建一个
The biometric system that uses multiple biometric traits promises higher identification accuracy than identification in either individual domain. To reach this goal, special attention should be paid to the strategies for combining voiceprint and face experts. We propose an improved weighted sum rule based on the scores difference (SDWS) between the genuine speaker class and the mistaken speaker class labeled by each classifier, and demonstrate that the performance of multibiometric system can be further improved by SDWS. The tests were conducted on a multi-modal database with 54 users We compare our approach with other existing methods and show that SDWS improved performance by about 7.8-13.3%, much better than the others.
[1] Gérard Chollet,et al. Comparing decision fusion paradigms using -NN based classifiers, decision trees and logistic regression in a multi-modal identity verification ap plication , 1999 .
[2] Josef Kittler,et al. An Experimental Comparison of Classifier Fusion Rules for Multimodal Personal Identity Verification Systems , 2002, Multiple Classifier Systems.
[3] Stefan Fischer,et al. Fusion of audio and video information for multi modal person authentication , 1997, Pattern Recognit. Lett..