<ul><li>Researchers at Cornell University are developing a wearable ring called SpellRing that combines deep learning AI and micro-sonar technologies to translate American Sign Language (ASL) fingerspelling gestures into text.</li><li>The ring features a quarter-sized 3D-printed casing with a microphone and speaker, worn around the thumb. It emits inaudible soundwaves that are detected as a miniature gyroscope measures hand motions. A deep-learning algorithm analyzes and translates the resultant sonar images into individual letters in real-time on a computer screen.</li><li>SpellRing's accuracy ranges from 82-92% after being trained with the help of 20 experienced and novice ASL signers spelling out over 20,000 words.</li><li>Although SpellRing is currently limited to fingerspelling, the researchers plan to expand its capabilities and develop devices for full ASL recognition in the future.</li></ul>

Wearable ring translates sign language into text

Discover more