摘要
为了提高语音验证技术的有效性,提出了一种基于共振峰合成、修改时长和调节韵律的随机语音验证码生成方法。该方法选择音素作为语音合成单元,基于规则在合成过程中设定随机语速参数,以及调整单元之间的连接规则来实现韵律的随机调整,使得语速和韵律具有不确定性和不可预测性,从而有效降低了自动语音识别技术(ASR)对语音码的识别率,增强了语音验证码的抗攻击性。合成的语音验证码的人耳识别率达到了90%左右,ASR的识别率为28.8%,主观平均判分(MOS)为4分,语音码的可懂度和清晰度达到了满意的效果。实验结果验证了所提方法的可行性。
In order to improve the effectiveness of speech verification technology, this paper proposed a method of speech validation codes based on formant synthesis, time scale modification and prosody regulation. This method chose phonemes as speech synthesis units and set parameters for speed regulations in the synthesis process based on rules, which adjusted the con- nection rules between units to achieve a random prosody regulation. Due to the uncertainty of speed and prosody, for speech val- idation codes, this method effectively reduced recognition rate of automatic speech recognition and enhanced resistance to at- tack. The recognition rate of synthesized speech validation codes was 90% for human ear, and 28.8% for automatic speech recognition software. The mean opinion score (MOS) was 4 points. Both intelligibility and articulation of the synthesis speech were satisfied. The experimental results confirm the practicality of the proposed method.
出处
《计算机应用研究》
CSCD
北大核心
2011年第7期2458-2461,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(61004112)
中国博士后科学基金资助项目(20080430750)
关键词
语音合成
验证码
共振峰合成
韵律调整
时长规整
speech synthesis
CAPTCHA
formant synthesis
prosody adjustment
time scale modification