Lea Schönherr (Ruhr University Bochum), Katharina Kohls (Ruhr University Bochum), Steffen Zeiler (Ruhr University Bochum), Thorsten Holz (Ruhr University Bochum), Dorothea Kolossa (Ruhr University Bochum)

Voice interfaces are becoming accepted widely as input methods for a diverse set of devices. This development is driven by rapid improvements in automatic speech recognition (ASR), which now performs on par with human listening in many tasks. These improvements base on an ongoing evolution of deep neural networks (DNNs) as the computational core of ASR. However, recent research results show that DNNs are vulnerable to adversarial perturbations, which allow attackers to force the transcription into a malicious output.

In this paper, we introduce a new type of adversarial examples based on *psychoacoustic hiding*. Our attack exploits the characteristics of DNN-based ASR systems, where we extend the original analysis procedure by an additional backpropagation step. We use this backpropagation to learn the degrees of freedom for the adversarial perturbation of the input signal, i.e., we apply a psychoacoustic model and manipulate the acoustic signal below the thresholds of human perception. To further minimize perceptibility of the perturbations, we use forced alignment to find the best fitting temporal alignment between the original audio sample and the malicious target transcription. These extensions allow us to embed an arbitrary audio input with a malicious voice command that is then transcribed by the ASR system, with the audio signal remaining barely distinguishable from the original signal. In an experimental evaluation, we attack the state-of-the-art speech recognition system *Kaldi* and determine the best performing parameter and analysis setup for different types of input. Our results show that we are successful in up to 98% of cases with a computational effort of fewer than two minutes for a ten-second audio file. Based on user studies, we found that none of our target transcriptions were audible to human listeners, who still understand the original speech content with unchanged accuracy.

View More Papers

Cleaning Up the Internet of Evil Things: Real-World Evidence...

Orcun Cetin (Delft University of Technology), Carlos Gañán (Delft University of Technology), Lisette Altena (Delft University of Technology), Takahiro Kasama (National Institute of Information and Communications Technology), Daisuke Inoue (National Institute of Information and Communications Technology), Kazuki Tamiya (Yokohama National University), Ying Tie (Yokohama National University), Katsunari Yoshioka (Yokohama National University), Michel van Eeten (Delft…

Read More

Enemy At the Gateways: Censorship-Resilient Proxy Distribution Using Game...

Milad Nasr (University of Massachusetts Amherst), Sadegh Farhang (Pennsylvania State University), Amir Houmansadr (University of Massachusetts Amherst), Jens Grossklags (Technical University of Munich)

Read More

Robust Performance Metrics for Authentication Systems

Shridatt Sugrim (Rutgers University), Can Liu (Rutgers University), Meghan McLean (Rutgers University), Janne Lindqvist (Rutgers University)

Read More

DIAT: Data Integrity Attestation for Resilient Collaboration of Autonomous...

Tigist Abera (Technische Universität Darmstadt), Raad Bahmani (Technische Universität Darmstadt), Ferdinand Brasser (Technische Universität Darmstadt), Ahmad Ibrahim (Technische Universität Darmstadt), Ahmad-Reza Sadeghi (Technische Universität Darmstadt), Matthias Schunter (Intel Labs)

Read More