Wavesdropper: Through-wall Word Detection of Human Speech via Commercial mmWave Devices

  title={Wavesdropper: Through-wall Word Detection of Human Speech via Commercial mmWave Devices},
  author={Chao Wang and Feng Lin and Zhongjie Ba and Fan Zhang and Wenyao Xu and Kui Ren},
  journal={Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.},
  • Chao WangFeng Lin Kui Ren
  • Published 2022
  • Computer Science
  • Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.
Most existing eavesdropping attacks leverage propagating sound waves for speech retrieval. However, soundproof materials are widely deployed in speech-sensitive scenes (e.g., a meeting room). In this paper, we reveal that human speech protected by an isolated room can be compromised by portable and commercial off-the-shelf mmWave devices. To achieve this goal, we develop Wavesdropper , a word detection system that utilizes a mmWave probe to sense the targeted speaker’s throat vibration and… 



WaveEar: Exploring a mmWave-based Noise-resistant Speech Sensing for Voice-User Interface

An in-depth study behind the rationale of voice generation and resulting vocal vibrations is performed, and the feasibility of employing radio-frequency signals, such as millimeter wave, for sensing the noise-resistant voice of an individual is investigated.

Acoustic Eavesdropping through Wireless Vibrometry

A new acoustic eavesdropping attack that can subvert such protectors using radio devices by inspecting the subtle disturbance it causes to the radio signals generated by an adversary or by its co-located WiFi transmitter is explored.

PitchIn: Eavesdropping via Intelligible Speech Reconstruction Using Non-acoustic Sensor Fusion

P PitchIn is presented to demonstrate the feasibility of speech reconstruction from non-acoustic sensor data collected offline across networked devices and how distributed TI-ADC can be used to achieve intelligibility by processing an interleaved signal composed of different sensors across networking devices.

VocalPrint: exploring a resilient and secure voice authentication via mmWave biometric interrogation

VocalPrint, a resilient mmWave interrogation system which directly captures and analyzes the vocal vibrations for user authentication, and demonstrates the resilience of the system against complex noise interference and spoof attacks of various threat levels.

Lamphone: Real-Time Passive Sound Recovery from Light Bulb Vibrations

This paper introduces "Lamphone," a novel side-channel attack for eavesdropping sound that can be used by eavesdroppers to recover human speech and singing and develops an algorithm to recover sound from the optical measurements obtained from the vibrations of a light bulb and captured by the electro-optical sensor.

Learning-based Practical Smartphone Eavesdropping with Built-in Accelerometer

A novel deep learning based system that learns to recognize and reconstruct speech information from the spectrogram representation of acceleration signals and employs adaptive optimization on deep neural networks with skip connections using robust and generalizable losses to achieve robust recognition and reconstruction performance.

WaveSpy: Remote and Through-wall Screen Attack via mmWave Sensing

This paper identifies and validate a new and practical side-channel attack for screen content via liquid crystal nematic state estimation using a low-cost radio-frequency sensor, and develops WaveSpy, an end-to-end portable through-wall screen attack system.

Speechless: Analyzing the Threat to Speech Privacy from Smartphone Motion Sensors

It seems that even machine-rendered speech may not be powerful enough to affect smartphone motion sensors through the aerial medium, although it may induce vibrations through a conductive surface that these sensors, especially accelerometer, could pick up if a relatively powerful speaker is used.

Hard Drive of Hearing: Disks that Eavesdrop with a Synthesized Microphone

This research demonstrates that the mechanical components in magnetic hard disk drives behave as microphones with sufficient precision to extract and parse human speech.

mmSense: Multi-Person Detection and Identification via mmWave Sensing

This paper proposes mmSense, a device-free multi-person detection and identification framework, which exploits the unique mmWave sensing features and proposes a novel approach to use humans' outline profile and vital signs to identify multiple people by using 60 GHz reflected signals of the human body.