AI Rings on Fingers Can Interpret Sign Language

Overview

A new study describes a set of electronic rings, wirelessly connected to an AI system, that translate multiple sign languages into text. Led by Ki Jun Yu of Yonsei University, the work uses seven rings with accelerometers and a deep-learning system to recognize signs without smart gloves or cameras. In testing with people who did not help train the system, it recognized 100 American Sign Language and 100 International Sign Language words with 88.3 and 88.5 percent accuracy.

Key Takeaways

Electronic rings wirelessly connected to an AI system translate multiple sign languages into text.
The work is led by Ki Jun Yu, an associate professor at Yonsei University in Seoul, Korea.
More than 300 different sign languages are used worldwide, and prior translation devices faced setbacks with cameras and smart gloves.
Researchers found seven fingers played major roles, so the system uses only seven rings, each with accelerometers.
The deep-learning system recognized signs from five people who did not take part in training.
The system reached 88.3 and 88.5 percent accuracy on 100 ASL and 100 International Sign Language words.

Stats & Key Facts

#More than 300 different sign languages used worldwide
#Seven rings used in the system
#88.3 percent accuracy on 100 American Sign Language words
#88.5 percent accuracy on 100 International Sign Language words
#Most previous systems were limited to fewer than 50 words
#Two people trained the system; five untrained people tested it

AI Rings on Fingers Can Interpret Sign Language

What the rings do

›Each electronic ring transmits its motion wirelessly to a processing device.
›Rings allow flexible sensor positioning to account for variations in people's hands.
›Wireless connections allow unrestricted hand motions.

Yu says the work is an important step toward making sign language translation systems more practical, lightweight, and usable in real-world environments.

Limits of earlier approaches

Prior translation projects faced repeated setbacks.

›Camera and computer-vision systems were limited to controlled settings and sensitive to lighting and interference.
›Smart gloves trapped heat and moisture, making prolonged use uncomfortable.
›Fixed glove sensors failed to account for hand size, finger length, and joint positions, reducing accuracy.
›Many wearable sensors required wired connections that hampered hand movement.

More than 300 different sign languages are used worldwide, and many projects aim to help communicate with people who do not know a sign language.

How the system is designed

›Researchers found seven fingers played major roles, so the system uses only seven rings to reduce hardware.
›Each ring uses accelerometers as inertial sensors to detect both stationary postures and hand movements.
›The team avoided bioelectric signals, which are highly specific to each person and need extensive calibration.

Yu notes Bluetooth Low Energy systems-on-chips have advanced enough that a wireless communication stack, power management circuit, and sensing module can fit on a flexible substrate small enough to wear as a ring.

Solving mechanical reliability

›Straight copper interconnects nearly broke under repeated bending.
›The team switched to serpentine-patterned interconnects that withstand repeated flexing.

Accuracy and remaining limits

A deep-learning system recognizes signs from hand movements.

›It identified signs from the two people used to train it and from five people who did not take part.
›It reached 88.3 percent accuracy on 100 common ASL words and 88.5 percent on 100 common International Sign Language words.
›Most previous attempts were limited to vocabularies of fewer than 50 words.

Dosik Hwang, a professor, cautions that 200 words is a meaningful advance over prior wireless systems but still a small fraction of a full sign language lexicon, which can contain thousands of signs.

Frequently Asked Questions

How accurate is the ring-based system?

In tests with five untrained people, it recognized 100 ASL words at 88.3 percent accuracy and 100 International Sign Language words at 88.5 percent accuracy.

Why use rings instead of gloves?

Rings allow flexible sensor positioning for different hand shapes and permit unrestricted hand motion, while smart gloves trapped heat and moisture and had fixed sensors that reduced accuracy.

How many rings does the system use and why?

Seven rings, because researchers found that seven fingers played major roles, which reduced the amount of hardware needed.

Who led the research?

Ki Jun Yu, an associate professor of electrical and electronic engineering at Yonsei University in Seoul, Korea.

What is the main limitation noted?

Professor Dosik Hwang cautions that 200 words, while an advance over prior wireless systems, is still a small fraction of a full sign language lexicon that can contain thousands of signs.

The study shows wireless AI-connected rings can translate two sign languages into text with about 88 percent accuracy, though the vocabulary remains far smaller than a full lexicon.