Can you Read This? How I Improved Text Readability Online for Visually Impaired
summary
Uses deep learning to detect illegible text over images and compares OCR engines tesseract and EasyOCR against a pixelwise text detector CRAFT model. Extracts character bounding boxes, trains a character-level readable/nonreadable classifier mixing subjective labels with objective contrast-standard labels, then thresholds the unreadable-character ratio to classify whole strings. Integrates into evinst site-scanning product and finds issues concentrated in travel, art/design and news sites, noting over 250 million people with moderate to severe visual impairment and ~36 million blind.
AI-written summary · about
chapters (6)
related talks
29mCan you Read This? (Or: how I Improved Text Readability on the Web for the Visually...
20mScene text access: a comparison of mobile OCR modalities for blind users
18mReadability & Design Building inclusive digital experiences
44mOCR Systems: Bringing Accessibility Where It’s Missing
48mTypography, Readability and Digital Accessibility
50mYou can't read this sentence
41mOCR in the Wild World of Social Media
51mReadability & Web: Let's build great inclusive projects