Audio Adversarial Examples: Attacks Using Vocal Masks

Autor:	Tay, Kai Yuan, Ng, Lynnette, Chua, Wei Han, Loke, Lucerne, Ye, Danqi, Chua, Melissa
Rok vydání:	2021
Předmět:	Computer Science - Sound Computer Science - Artificial Intelligence Electrical Engineering and Systems Science - Audio and Speech Processing
Druh dokumentu:	Working Paper
Popis:	We construct audio adversarial examples on automatic Speech-To-Text systems . Given any audio waveform, we produce an another by overlaying an audio vocal mask generated from the original audio. We apply our audio adversarial attack to five SOTA STT systems: DeepSpeech, Julius, Kaldi, wav2letter@anywhere and CMUSphinx. In addition, we engaged human annotators to transcribe the adversarial audio. Our experiments show that these adversarial examples fool State-Of-The-Art Speech-To-Text systems, yet humans are able to consistently pick out the speech. The feasibility of this attack introduces a new domain to study machine and human perception of speech. Comment: 9 pages, 1 figure, 2 tables. Submitted to COLING2020
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2102.02417 Zobrazit plný text záznamu View this record from Arxiv