A team of six computer science and computer information technology students won the 2017 CSUN Accessibility Competition, held February 3-5 on the CSUN campus. Entrants were challenged to develop a CAPTCHA alternative for elderly mobile users. CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) is a verification technique that has become essential for defending against malicious bots in Internet operations. Most versions rely on keyboard input and language interpretation; for example, the most common CAPTCHA presents users with random, distorted characters and requires them to type them into a box.
On mobile devices, CAPTCHA can be challenging, especially for elderly users with physical impairments. Because of the small display area, lower screen resolution, display color combinations, limited processing power and virtual keyboards typical of smartphones and tablets, recognizing the images and inputting responses can be difficult. Consequently, CAPTCHAs are not accessible to all users, some of whom won’t be able to carry out Internet or mobile tasks as a result. Visually impaired users, for example, will find it very difficult or even impossible to complete an image-based CAPTCHA.
Participants in the competition were required to develop their solution for older adults who fell into one of the following groups: deaf or hearing impaired; blind or visually impaired; or with other sensory impairments, such as fine motor impairments or memory, learning and cognitive impairments. A workshop on January 28 provided students with the basics of mobile development and machine learning, and four CECS teams competed in the weekend event.
The solution from the winning team (made up of Ayrin Golestanian, Beulah Preethi Vallur, Mohammad Laknahour, Carlos Sandoval, Kyle Astadillo and Yul Puma) made the mobile human verification process accessible by enabling users with visual impairments and limited fine motor skills to navigate the process via prompts from the phone. The team plans to continue refining their solution and to compete against other institutions in the 19th International ACM SIGACCESS Conference on Computers and later this year in Baltimore.
In addition to a talented team’s savvy programming, support from industry helped make the victory possible, including sponsorship from Northrop Grumman and mentoring by Casey Smith, an accessibility specialist from CapitalOne. The competition was organized by Li Liu, an assistant professor in the computer science department.