论文标题
消失的命令：用声音掩盖对自动语音识别系统的欺骗攻击
Disappeared Command: Spoofing Attack On Automatic Speech Recognition Systems with Sound Masking
论文作者
论文摘要
深度学习技术的发展极大地促进了自动语音识别（ASR）技术的性能提高，该技术证明了在许多任务中与人类听力相当的能力。语音接口正在变得越来越广泛地用作许多应用程序和智能设备的输入。但是，现有的研究表明，DNN很容易受到轻微干扰的干扰，并且会出现错误的识别，这对于由语音控制的智能语音应用非常危险。
The development of deep learning technology has greatly promoted the performance improvement of automatic speech recognition (ASR) technology, which has demonstrated an ability comparable to human hearing in many tasks. Voice interfaces are becoming more and more widely used as input for many applications and smart devices. However, existing research has shown that DNN is easily disturbed by slight disturbances and makes false recognition, which is extremely dangerous for intelligent voice applications controlled by voice.
