Software reads online content aloud and printers generate Braille text, but there hasn't been a fast and easy way to create recognizable images for the blind. Now, computer scientists in Arizona are generating social networking profile pictures the blind can "see."
"The face image — that's very important for people in their social life, emotional life," said Baoxin Li, an associate professor of computer science at Arizona State University who is leading the software work.
Li said the idea was inspired by a blind ASU researcher who wished she could access more graphical information. Making all digital graphics accessible to the blind would have been an overwhelming challenge, so Li and his colleagues focused on profile pictures. They had to find the right balance of information so the person would be recognizable.
"We convert the photo in such a way so the major facial landmarks are nicely kept — that's very important because we can't render all the features into tactile form," Li said. "That would be too disorienting."
Instead, an algorithm pares down crucial facial information without oversimplifying it. Their software allows a blind user to take a photo of a face, put it into a computer application, and automatically generate a new printable image. The image comes out of a special tactile printer with raised lines along the facial features.
"At the moment it's within one minute or so, but we can further optimize the software to do it faster," Li said.
Tactile printers are usually found at centers that assist the blind, and institutions such as the Center for Cognitive Ubiquitous Computing at ASU. However, Li said that even the least expensive ones cost several thousand dollars. In the future, he expects the software will work with paperless tactile displays that are in development.
Their automated approach was described last year in the journal IEEE Transactions on Multimedia. This week, Li demonstrated the software at the International Conference on Intelligent User Interfaces in Palo Alto, Calif.
Other technology exists for creating tactile images, Li said, but it's designed to help sighted professionals with the time-consuming process of making intricate images for the blind. The ASU software is stable to the point where the scientists are talking with software producers about bringing it to market.
Beyond profile images, the scientists would like to create software that can generate tactile images from online mapping sites.
John Gardner is a former Oregon State University physics professor who lost his vision in 1988. Frustrated by a lack of access to information, he founded the assistive technology company ViewPlus Technologies in Corvallis, Ore. ViewPlus developed the tactile printer that Li uses, he said.
"But we never had the software to make a nose feel like a nose and an eye feel like an eye," Gardner said. "It's a tour de force that he can analyze a face and make it feel like a face." He added that he'd like Li's software to render the Mona Lisa.
At the demonstration this week in California, attendees were invited to have their photos taken and receive tactile versions. Some sighted visitors called the printouts works of art, Li said. "They asked me to put my signature on their copies."