arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2203.03706cs.SDcs.LGeess.AS
收藏

AI合成印地语语音的检测

Detection of AI Synthesized Hindi Speech

Dhirubhai Ambani Institute of Information and Communication Technology(迪鲁巴伊·安巴尼信息与通信技术学院);Indian Institute of Technology Jammu(印度理工学院查谟分校)

Karan Bhatia, Ansh Agrawal, Priyanka Singh, Arun Kumar Singh

摘要:

生成式人工语音模型的最新进展使得生成高度逼真的语音信号成为可能。起初,获得这些人工合成的信号(如语音克隆或深度伪造)似乎令人兴奋,但如果不加以控制,它可能会导致我们走向数字反乌托邦。音频取证的主要焦点之一是验证语音的真实性。虽然针对英语语音提出了一些解决方案,但对合成印地语语音的检测还没有得到太多的关注。在这里,我们提出了一种方法,从一个实际的人类语音人工智能合成印地语语音的歧视。我们已经利用双相干相位、双相干幅度、梅尔频率倒谱系数(MFCC)、Delta倒谱系数和Delta平方倒谱系数作为机器学习模型的鉴别特征。此外,我们将研究扩展到使用深度神经网络进行广泛的实验,特别是VGG 16和自制的CNN作为架构模型。我们获得了99.83%的准确性与VGG 16和99.99%的自制CNN模型。

英文摘要:

The recent advancements in generative artificial speech models have made possible the generation of highly realistic speech signals. At first, it seems exciting to obtain these artificially synthesized signals such as speech clones or deep fakes but if left unchecked, it may lead us to digital dystopia. One of the primary focus in audio forensics is validating the authenticity of a speech. Though some solutions are proposed for English speeches but the detection of synthetic Hindi speeches have not gained much attention. Here, we propose an approach for discrimination of AI synthesized Hindi speech from an actual human speech. We have exploited the Bicoherence Phase, Bicoherence Magnitude, Mel Frequency Cepstral Coefficient (MFCC), Delta Cepstral, and Delta Square Cepstral as the discriminating features for machine learning models. Also, we extend the study to using deep neural networks for extensive experiments, specifically VGG16 and homemade CNN as the architecture models. We obtained an accuracy of 99.83% with VGG16 and 99.99% with homemade CNN models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑