SearchAccess: Advancing Accessibility in Android App Design Through A Deep Learning-Powered GUI-Based Search Engine


  • SOPHIA LIN Thomas Jefferson High School for Science and Technology, Alexandria, VA
  • Arun Krishnavajjala Department of Computer Science, George Mason University, Fairfax, VA
  • Kevin Moran Department of Computer Science, George Mason University, Fairfax, VA and Department of Computer Science, University of Central Florida, Orlando, FL



Accessibility issues persistently limit access and usability of apps for individuals with diverse needs. Developers often struggle to determine the accessibility of their designs, encountering challenges in identifying and addressing potential barriers, especially with numerous disabilities such as motor impairment and visual impairment. Many tools and indicators have been created to test the implementation for accessibility issues. However, there is little support in the design process to ensure that software engineers have user-friendly designs prior to implementation and testing. We present a Graphical User Interface (GUI)-based search engine that allows developers to compare a screenshot of their preliminary designs of their Android app with a set of similar screenshots in accessibility scores using a variety of tests. This allows developers to ensure their designs are accessible before spending significant time implementing and testing them. By offering developers a tool to find similar yet more accessible designs, we hope to guide the design of accessible apps before implementation and testing, improving accessibility in apps with less effort and time spent by the developer.





College of Engineering and Computing: Department of Computer Science