TY - GEN
T1 - Partially Fake Audio Detection Based on MOSNet with Pretraining Models
AU - Liu, Hanyue
AU - Zhang, Jianqian
AU - Wang, Jing
AU - Liu, Miao
AU - Xu, Liang
AU - Sun, Yi
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - With the rapid development of speech synthesis and voice conversion related technologies, many potential risks have been brought to people's information security and privacy. Therefore, it is important to build techniques to identify manipulated regions in audios. In this paper, we propose a novel partially fake audio detection system based on MOSNet, a speech quality assessment network, and pretraining models. Comparisions between features extracted by pretraining models and Mel-spectrogram are made. Experimental results show that the proposed system combining MOSNet and XLS-R-300m pretraining model has the best performance on both evaluation set and test set, and has good generalization ability. The final score of the proposed system on test set is 5.97% higher than that of the baseline system based on RawNet.
AB - With the rapid development of speech synthesis and voice conversion related technologies, many potential risks have been brought to people's information security and privacy. Therefore, it is important to build techniques to identify manipulated regions in audios. In this paper, we propose a novel partially fake audio detection system based on MOSNet, a speech quality assessment network, and pretraining models. Comparisions between features extracted by pretraining models and Mel-spectrogram are made. Experimental results show that the proposed system combining MOSNet and XLS-R-300m pretraining model has the best performance on both evaluation set and test set, and has good generalization ability. The final score of the proposed system on test set is 5.97% higher than that of the baseline system based on RawNet.
KW - deep learning
KW - manipulation region location
KW - partially fake audio detection
KW - pretraining models
UR - http://www.scopus.com/inward/record.url?scp=85194135350&partnerID=8YFLogxK
U2 - 10.1109/ACAIT60137.2023.10528491
DO - 10.1109/ACAIT60137.2023.10528491
M3 - Conference contribution
AN - SCOPUS:85194135350
T3 - Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023
SP - 899
EP - 903
BT - Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023
Y2 - 10 November 2023 through 12 November 2023
ER -