What it does
OTON GLASS is a set of smart glasses designed to help dyslexic people to read words. The glasses have two small cameras and an earpiece. The camera captures pictures of words that user wants to read and reads out the words for the user via the earpiece.
I was determined to develop OTON GLASS through my father’s dyslexia experience. In 2012, my father had a brain tumor, and developed dyslexia after his operation – catalyst for OTON GLASS. Fortunately, he recovered fully after rehabilitation. However, many people have congenital dyslexia regardless of their symptoms. First, I interviewed my father and observed his behavior and identified problems. During the ideation phase, I did idea sketching and video prototyping. I formed a team of engineers and product designers who invented an active prototype. I saw the necessity to develop OTON GLASS after receiving feedbacks from my father and 5 other dyslexic people.
How it works
OTON GLASS utilizes Raspberry Pi (RPi) as its computer. The glass’s sides have two cameras and an earphone, which connects to the RPi. The lens is half mirror and reflects the user’s eye and the inside camera tracks the eye movements. One camera detects blinks as a trigger while the other one captures words. The RPi processes these images by OCR and detects the words in the images. The artificial voice technology changes words to sounds, which the user can then understand it’s meaning. This OTON GLASS prototype is constantly being updated. The current OCR (including API) - designed by an outside academic group is capable of changing words to sound; however, the OCR cannot recognize ever character. Therefore, we considering to make OTON GLASS sort the desired words to be read, and if the RPI can’t recognize them, it will send to a remote supporter that can decipher the words. In this update UX scenario, OTON GLASS realizes the scenario that the user desires.
I interviewed my father and observed his behavior. During the interview, he told me about having difficulty reading in his everyday life. Through the practices of cultural anthropology, I settled on the theme “a device that changes written words to sounds”. I generated ideas and researched technologies. I made a video prototype and communicated with engineers and product designers for collaboration. We developed the first working prototype that could change words to sounds in 15 seconds. With this prototype, we obtained feedbacks from my father’s usage. We then developed a new prototype for other dyslexic people. This second prototype was modeled and designed via CAD and printed with a 3D printer. With this improved prototype, we incorporated additional eye tracking functions that enabled the feedback speed to be shortened to 3 seconds. We obtained feedback from 5 dyslexic people on the new prototype. Some examples of the feedback included: 1) want to have a double check feature when reading; 2) want to use when user has to read quickly. With this valuable feedback from user testing, we were able to discover some of the situations where the prototype would be used, and determined our next set of tasks. Presently, we are continuously developing prototypes to improve OTON GLASS.
How it is different
There exist similar products in the world, but they are not currently commercialized yet. They require a breakthrough of technology and trial-and-error on how to deploy smart glasses. The originality of OTON GLASS consists of two aspects, technology and deployment. First, in the technology realm, startups such as Orcam Inc. and Hours Technology Inc. are currently developing smart glasses for blind people. They mainly develop powerful OCR for the English (Alphabet) using machine learning techniques. On the other hand, OTON GLASS focuses on Japanese character recognition as its unique aspect. OTON GLASS aims to solve the user’s problems by becoming a hybrid (human-to-computer) recognizer and not approaching the problem using OCR Technology. Secondly, in terms deployment, OTON GLASS is all in one that combines camera-to-glasses – meaning they look like normal glasses. This capture trigger based on human’s behavior is natural interaction for people.
Our goal is to give OTON GLASS to the people who are in need. We believe in developing prototypes with many people from different fields, and it will expand our community and get enough funds for the further R＆D. Now, our team consists of Computer Vision and Machine Learning researcher, hardware engineer, product designer, and a 3D modeller. We will repeat developing prototypes and doing user tests in the short-term. This prototype will grow and become a “product”. When we complete and develop a prototype that the user can use, we will do cloudfunding and extend this concept. Finally, we will design a service with NPOs and hospitals for the users can get it easily.
Social architects prize at NTT Docomo incubation program’s Demoday, 2014 Mayor’s Prize at Cyber glasses AR app contest, 2014 Grand Prix at NTT Docomo developer application contest, 2014