Adversarial Attacks Against Automatic Speech Recognition Systems via Psychoacoustic Hiding
Preprint
Abstract
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 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 the 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.Keywords
All Related Versions
This publication has 12 references indexed in Scilit:
- PRADA: Protecting Against DNN Model Stealing Attacks2019
- Understanding adversarial training: Increasing local stability of supervised models through robust optimizationNeurocomputing, 2018
- Building Competitive Direct Acoustics-to-Word Models for English Conversational Speech Recognition2018
- Toward Human Parity in Conversational Speech RecognitionIEEE/ACM Transactions on Audio, Speech, and Language Processing, 2017
- Improved Gender Independent Speaker Recognition Using Convolutional Neural Network Based Bottleneck Features2017
- Tied Variational Autoencoder Backends for i-Vector Speaker Recognition2017
- Improved wavelet-based watermarking through pixel-wise maskingIEEE Transactions on Image Processing, 2001
- A guided tour to approximate string matchingACM Computing Surveys, 2001
- Audio watermarking for copyright protection ofdigital audio dataElectronics Letters, 2001
- Perceptual watermarks for digital images and videoProceedings of the IEEE, 1999