Scientists have developed glasses that “read lips” without using a camera


Cornell University researchers have developed a new technology that enables silent communication through sonar glasses, allowing wearers to perform a variety of tasks without physical input. The glasses use tiny speakers and microphones to read words silently uttered by the wearer, making it an improvement on a previous project that used a wireless headset.

How the Technology Works

The glasses employ a silent speech recognition interface called EchoSpeech, which uses sonar to sense mouth movements, and a deep learning algorithm to analyze echo signatures in real-time. With an accuracy rate of about 95%, the system can identify the words the wearer is silently saying.

Potential Applications

The technology has promising prospects, including the ability to assist people with speech impairments, enabling them to silently input conversations into a speech synthesizer and speak out loud. The glasses can also be used to control music playback in a quiet library or dictate messages at a noisy concert.

Additionally, the technology is small, low-power, and does not pose privacy concerns because no data leaves the user’s phone. It is easy to wear and more practical than other silent speech recognition technologies available.

Training and Functionality

To learn the user’s speech patterns, the system only requires a few minutes of training data. After training, it can receive and send sound waves to the user, analize mouth movements, and analyze the echo signatures using deep learning algorithms. The system can recognize 31 isolated commands and a string of consecutive numbers with an error rate of less than 10%.

Battery Life and Wireless Connectivity

The current version of the system offers approximately 10 hours of battery life and can communicate wirelessly with a user’s smartphone via Bluetooth. The smartphone processes and predicts all the data and transmits the results to a few “action keys” that enable the glasses to play music, interact with smart devices, or activate voice assistants.

Commercialization Potential

Cornell University’s Intelligent Computer Interface Future Interaction (SciFi) Lab is exploring the possibility of commercializing the technology using a Cornell grant program. With further development and commercialization, it may become an essential tool for individuals with disabilities or those seeking to improve their communication capabilities.