摘要
围绕声音伪造与防伪检测技术,首先梳理声音伪造与防伪检测技术的发展脉络,其次分析伪造声音的语音合成和语音转换技术的原理,以及声音防伪检测技术的原理,最后对声音防伪检测技术提出构建大规模真伪声音数据库、进一步挖掘可辨识真伪声音的特征、优化防伪检测模型三点发展建议。
This paper focus on audio spoofing and anti-spoofing detection.Firstly,development of audio spoofing and audio spoofing detection is introduced.Then we analyze the principle of audio spoofing(speech synthesis and voice conversion)and audio spoofing detection.Finally,three suggestions are put forward,aiming at improving the performance of anti-spoofing detection:building big-scale database of real audio and spoofing audio,extract features for audio spoofing detection and training optimization model.
出处
《信息技术与标准化》
2020年第1期54-58,共5页
Information Technology & Standardization
关键词
声音伪造
声音防伪检测
语音合成
语音转换
audio spoofing
audio spoofing detection
speech synthesis
voice conversion