I'm in love with the shape
summary
Live demo shows the Shape Detection API (barcode, face, text) used from the browser to augment alt text: barcode is default in Chrome 83, face and text are behind flags and require experimental web platform features. CodePen uses cross-origin anonymous images; face detector returns bounding boxes and landmarks but on macOS requires fastMode: true to find faces; text detector sometimes returns empty rawValue on macOS but works on Windows. Result: alt becomes image contains 1 faces plus joined detected text.
AI-written summary · about
chapters (7)
- 0:00Start
- 1:24Picture Is Worth a Thousand Words
- 5:14Shape Detection Api
- 8:46Create a Face Detector
- 9:15Face Detector
- 18:08Text Detector
- 23:26Barcode Detection
related talks
32mSelf-Portraits in JavaScript
19mNew Tools for Old Problems
49mReal Time Face Detection through a Webcam Using AI
21mBring AI-Based Search to Your Web App
22mFAULT TEXT: The Adversities in Automated Alternative Text with Henri Helvetica
01h 09mThe wonderful world of new Web APIs
22mHI and AI
50mScrutinizing @font-face
