This article discusses a study conducted by Guo et al. at Michigan State University, which focuses on preventing someone from recording a person’s voice. The study explores the use of ultrasound signals to cancel out a person’s voice in a recording device. The researchers utilize micro-electro-mechanical system (MEMS) microphones commonly found in smartphones and other recording devices. They develop a neural network called Neurally Enhanced Cancellation (NEC) system, which is trained on voice samples of the person whose voice needs to be cancelled. The ultrasound signal is then sent to the target recording device, effectively removing the voice audio signal from the final recording. The article highlights that this system can only jam a specific microphone and does not affect hidden microphones in a room. It also mentions the legal implications of general microphone jamming and how the NEC system overcomes those issues.
| Signal | Change | 10y horizon | Driving force |
|---|---|---|---|
| Using ultrasound to cancel out voices in recordings | From recording voices to preventing recording | Increased privacy in voice recordings | Desire for privacy and protection of personal data |