Steganalysis of adaptive multi-rate speech using statistical characteristics of pulse pairs

详细信息    查看全文

• 作者：Hui Tian^a ; ^{htian@hqu.edu.cn} ; Yanpeng Wu^a ; Chin-Chen Chang^b ; Yongfeng Huang^c ; Yonghong Chen^a ; Tian Wang^a ; Yiqiao Cai^a ; Jin Liu^a
• 关键词：Steganalysis ; Steganography ; Adaptive multi-rate speech ; Pulse-pair statistical characteristics ; Adaptive boosting ; Support vector machine
• 刊名：Signal Processing
• 出版年：2017
• 出版时间：May 2017
• 年：2017
• 卷：134
• 期：Complete
• 页码：9-22
• 全文大小：4476 K
• 卷排序：134

文摘

We characterize AMR speech exploiting the statistical properties of pulse pairs. Three types of steganalysis features are presented. We present an SVM-based steganalysis scheme with feature selection using AdaBoost. Our method significantly outperforms the state-of-the-art steganalysis methods. Our method can be applied in steganalysis scenarios for real-time speech streams.