A Cornell University lab has developed eyeglasses that use AI and acoustic-sensing technology to recognize commands from the wearer without requiring them to speak. The EchoSpeech glasses can "understand" up to 31 commands based on only mouth and lip movements.
- The
commands do not need to be vocalized, which could give patients that
can't speak "their voices back," according to Cornell doctoral student
Ruidong Zhang, lead author of the research paper on the glasses.
- EchoSpeech is a "silent-speech recognition interface" that uses microphones, small speakers, and an AI-powered sonar system.
- Soundwaves emitted from a person's mouth movements are received and processed using a deep learning algorithm.
- The algorithm "analyzes these echo profiles in real-time," with a reported accuracy of around 95%.
- The
system was created by researchers at Cornell's Smart Computer
Interfaces for Future Interactions (SciFi) Lab, who are looking into
commercializing EchoSpeech's technology.
- "It's small,
low-power, and privacy-sensitive, which are all important features for
deploying new, wearable technologies in the real world," SciFi Lab
director Cheng Zhang said.
