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 2102.00313cs.SDcs.LGeess.AS
收藏

防御对抗性音频攻击的大脑皮层特征

Cortical Features for Defense Against Adversarial Audio Attacks

University of Maryland, College Park(马里兰大学帕克分校);Johns Hopkins University(约翰斯·霍普金斯大学)

Ilya Kavalerov, Ruijie Zheng, Wojciech Czaja, Rama Chellappa

摘要:

我们建议使用听觉皮层的计算模型作为对音频的对抗攻击的防御。我们将几种基于白盒迭代优化的对抗攻击应用于Amazon Alexa硬件网络的实现,以及具有集成皮质表示的该网络的修改版本,并表明皮质特征有助于防御通用对抗示例。在相同的失真水平下,针对皮层网络发现的对抗性噪声对于通用音频攻击总是不太有效。我们在https://github.com/ilyakava/py3fst上公开发布我们的代码。

英文摘要:

We propose using a computational model of the auditory cortex as a defense against adversarial attacks on audio. We apply several white-box iterative optimization-based adversarial attacks to an implementation of Amazon Alexa's HW network, and a modified version of this network with an integrated cortical representation, and show that the cortical features help defend against universal adversarial examples. At the same level of distortion, the adversarial noises found for the cortical network are always less effective for universal audio attacks. We make our code publicly available at https://github.com/ilyakava/py3fst.

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。

↑