In recent years, the performance of Automatic Speaker Verification (ASV) systems has been improved significantly. However, they are still affected by different kind of spoofing attacks. In this paper, we propose a method that fused different phase features and amplitude features to detect replay attacks. We propose the mel-scale relative phase feature and apply source-filter vocal tract feature in phase domain for replay attacks detection. These two phase-based features are combined to get complementary information. In addition to these phase haracteristics, constant Q cepstral coefficients (CQCCs) are used. The proposed methods are evaluated using the ASVspoof 2017 challenge database and Gaussian mixture model was used as the back-end model. The proposed approach achieved 55.6% relative error reduction rate than the conventional magnitude-based feature.
Cite as: Li, D., Wang, L., Dang, J., Liu, M., Oo, Z., Nakagawa, S., Guan, H., Li, X. (2018) Multiple Phase Information Combination for Replay Attacks Detection. Proc. Interspeech 2018, 656-660, doi: 10.21437/Interspeech.2018-2001
@inproceedings{li18n_interspeech,
author={Dongbo Li and Longbiao Wang and Jianwu Dang and Meng Liu and Zeyan Oo and Seiichi Nakagawa and Haotian Guan and Xiangang Li},
title={{Multiple Phase Information Combination for Replay Attacks Detection}},
year=2018,
booktitle={Proc. Interspeech 2018},
pages={656--660},
doi={10.21437/Interspeech.2018-2001}
}