Isolated and Ensemble Audio Preprocessing Methods for Detecting Adversarial Examples against Automatic Speech Recognition
This addresses a security threat in limited-vocabulary speech recognition systems, such as those used in telephony, by improving detection of adversarial attacks, though it is incremental as it builds on existing preprocessing techniques.
The paper tackles the problem of detecting adversarial examples in automatic speech recognition, specifically against the Speech Commands Model, by exploring audio preprocessing methods. It reports that a combined defense using compression, speech coding, filtering, and audio panning achieved 93.5% precision and 91.2% recall in detecting adversarial examples.
An adversarial attack is an exploitative process in which minute alterations are made to natural inputs, causing the inputs to be misclassified by neural models. In the field of speech recognition, this has become an issue of increasing significance. Although adversarial attacks were originally introduced in computer vision, they have since infiltrated the realm of speech recognition. In 2017, a genetic attack was shown to be quite potent against the Speech Commands Model. Limited-vocabulary speech classifiers, such as the Speech Commands Model, are used in a variety of applications, particularly in telephony; as such, adversarial examples produced by this attack pose as a major security threat. This paper explores various methods of detecting these adversarial examples with combinations of audio preprocessing. One particular combined defense incorporating compressions, speech coding, filtering, and audio panning was shown to be quite effective against the attack on the Speech Commands Model, detecting audio adversarial examples with 93.5% precision and 91.2% recall.