South Korean POSTECH designs neckband converting silent mouthing into speech
Technology
Scientists at POSTECH in South Korea have designed a smart neckband that turns silent mouth movements into actual speech.
It picks up tiny muscle and skin shifts on your neck as you mouth words, then uses AI to recreate your voice, even if you don't say anything out loud.
Training takes less than 10 minutes, and it nailed an 85.8% accuracy rate with a test vocabulary of 26 words.
Prototype upgrades aim to broaden applications
This tech could be a game-changer for loud workplaces like construction sites or for people who've lost their voice.
Originally meant for medical use, the neckband is still being upgraded.
Future versions should handle more words and work better while users are moving, making it even more versatile across different fields.