How Private is Low-Frequency Speech Audio in the Wild? An Analysis of Verbal Intelligibility by Humans and Machines

TL;DR

This study evaluates whether low-frequency speech audio recordings effectively protect verbal privacy in real-world social settings by analyzing human and machine speech recognition performance under various noise conditions.

Contribution

It provides a comprehensive analysis of the privacy-preserving potential of low-frequency audio, including simulated privacy attacks and the trade-offs with speech detection accuracy.

Findings

Low-frequency audio reduces automatic speech recognition accuracy.

Humans can still understand low-frequency speech in certain conditions.

Trade-offs exist between privacy and social behavior analysis accuracy.

Abstract

Low-frequency audio has been proposed as a promising privacy-preserving modality to study social dynamics in real-world settings. To this end, researchers have developed wearable devices that can record audio at frequencies as low as 1250 Hz to mitigate the automatic extraction of the verbal content of speech that may contain private details. This paper investigates the validity of this hypothesis, examining the degree to which low-frequency speech ensures verbal privacy. It includes simulating a potential privacy attack in various noise environments. Further, it explores the trade-off between the performance of voice activity detection, which is fundamental for understanding social behavior, and privacy-preservation. The evaluation incorporates subjective human intelligibility and automatic speech recognition performance, comprehensively analyzing the delicate balance between effective social behavior analysis and preserving verbal privacy.