Glove translates sign language into speech via an app
A glove has been developed that translates sign language into speech in real time.
Scientists from the University of California, LA, have created an “inexpensive, flexible and highly durable” device that is capable of turning American sign language into English speech via an app.
Read more: Deaf mother designs mask that enables people to lip read
The glove contains thin, stretchable sensors along the length of each of the fingers. These pick up hand motions and finger placements that stand for letters, numbers, words and phrases.
The movements are then turned into electrical signals that are sent to a small circuit board on the wrist.
Read more: Shouting during pregnancy may impact your baby’s hearing
Finally, the board transmits those signals wirelessly to a smartphone, where they get translated at the rate of around one word a second.
Although encouraging, the glove was tested on just four deaf patients. It is also unclear when it may become publicly available.
Around 37.5 million people (15%) aged 18 or over in the US “report some trouble hearing”.
In the UK, 11 million people are thought to have hearing loss, making it the second most common disability.
American sign language is “the primary language of many North Americans who are deaf and hard of hearing”.
It is defined as “a complete, natural language that has the same linguistic properties as spoken languages, with grammar that differs from English”.
Read more: Hearing test at birth aid ‘may aid early autism diagnosis’
Speaking of the glove, lead author Dr Jun Chen said: “Our hope is this opens up an easy way for people who use sign language to communicate directly with non-signers without needing someone else to translate for them.
“In addition, we hope it can help more people learn sign language themselves.”
When developing the glove, the scientists added adhesive sensors to the four testers’ faces, between their eyebrows and on one side of their mouth.
This enabled them to capture facial expressions that are a part of American sign language.
The participants repeated 660 sign language gestures 15 times. A custom machine-learning algorithm then turned these into letters, numbers, words and phrases.
When put to the test, the glove demonstrated a recognition rate of up to 98% in a time of less than one second, according to results published in the journal Nature Electronics.
Similar systems were previously found to be bulky and heavy, making them uncomfortable to wear, according to the scientists.
Their glove is made of lightweight and inexpensive – but still long-lasting and stretchable – polymers, they added. The electronic sensors are also said to be flexible and relatively cheap.
Although unclear how much it may one day cost, the scientists claim the glove was constructed from components that came in at less than $50 (£40.55) “at a laboratory scale”.
