Blind and Human: Exploring More Usable Audio CAPTCHA Designs
summary
A three-session online study with 67 visually impaired participants (33 completed all sessions) creates 39 audio CAPTCHAs using Audacity, served via an HTML/jQuery/PHP testbed on Heroku and vetted over Zoom. Four new designs—math, character, pauses, categories—beat the control (43% accuracy, 53.6 s): math 89% (31.7 s), character 87% (32.7 s), pauses 76%, categories 70% (fastest). Security testing contrasts vulnerability to random guessing versus open‑source NLP attacks, and user ratings favor character for usability while math yields highest accuracy.
AI-written summary · about
chapters (6)
related talks
26mDisruptive Design: Harmful Patterns and Bad Practice
25mSecurity and Accessibility can be friends!
05mWebAlly: Making Visual Task-based CAPTCHAs Transferable for People with Visual
13mIterative Design of An Accessible Crypto Wallet for Blind Users
01h 01mPracticing Accessibility Research with Confidence
52mAccessibility: Coding and Beyond
18mMethods for Evaluation of Imperfect Captioning Tools by Deaf or Hard-of-Hearing Users at ...
38mGuidelines on Designing for Accessibility in UX Design