It works in under 0.3 seconds

StethoSpeech works in under 0.3 seconds, even if you're moving around.
You can personalize the voice for accent or gender by providing just a few hours of murmuring data.
Unlike older tech that needs invasive sensors, this system is wireless and noninvasive: just a regular stethoscope does the trick.

The team trained StethoSpeech with data from busy places

The team trained StethoSpeech with data from busy places like offices and concerts so it works reliably even when things get loud.
Its smart pipeline mimics how humans learn to speak, making it both practical and pretty innovative for assistive tech.