Mobile phone identification from speech recordings using Weighted Support Vector Machine

Abstract

In this paper, we propose a mobile phone identifier called Weighted Support Vector Machine with Weighted Majority Voting (WSVM-WMV) for a closed-set mobile phone identification task. The proposed WSVM-WMV can be regarded as a generalization of the traditional SVM identifier. On using Mel-frequency Cepstral Coefficients (MFCC) and Linear-frequency Cepstral Coefficients (LFCC) as the feature vectors, the proposed identifier can improve the identification accuracy from 92.42% to 97.86% and from 90.44% to 98.33% respectively, as compared with the traditional SVM identifier in identifying a set of 21 mobile phones.