针对语音变换的语音篡改检测

针对使用语音变换技术的语音篡改,提出一种自动检测方法。在分析语音变换基本模型和变换语音失真的基础上,提取语音信号的声道参数以及相关的信号统计量,并通过支持向量机递归特征消除法,选择出对语音变换比较敏感的特征作为分类特征,使用支持向量机进行语音变换检测和变换语音的说话人性别判别。对于一种语音变换软件的实验结果表明,该方法具有较高的检测准确率,其中语音变换检测的平均准确率为94.90%,变换语音的说话人性别判别平均准确率为92.09%。

Speech Tampering Detection for Voice Transformation

An automatic speech tampering detection method for voice transformation is proposed.Based on the analysis of the basic model of voice transformation and distortions introduced by it,vocal tract parameters and statistics of speech signals are extracted.Sensitive features are selected as classifying features by support vector machine recursive feature elimination method.Support vector machine classifiers are applied to get the decision of voice transformation detection and speaker gender of transformed speech.Experimental results on a kind of voice transformation software show that the proposed method has high detection accuracy,where the average accuracy is 94.90% for voice transformation and 92.09% for speaker gender detection of transformed speech.