Speech Recognition as a Progressive Enhancement
summary
Proposes VoiceHTML as a protocol-like idea and breaks it into four phases including a yearly International Screen Reader day, enhanced screenreaders and a personal proxy called Robbie the Reader; compares booking an airline ticket with today’s screenreader workflow versus a VoiceHTML-enhanced site; argues no new APIs are required — use ARIA, the accessibility tree, Web Speech API, WebSockets and JavaScript — and says the real gap is a web-like trust model for voice rather than new tech.
AI-written summary · about
chapters (28)
- 0:00Introduction
- 1:55Background
- 3:49The Web
- 8:12iPhone 2007
- 11:58Infoseek
- 12:53Links
- 13:33Phones
- 13:59World Wide Web
- 15:13Money
- 16:24Trust Cloud
- 17:10Choice
- 18:13Trust Model
- 19:26User Agent
- 19:58Apple App Store
- 20:42Voice
- 21:53New API
- 22:18VoiceHTML
- 22:49International Screenreader Day
- 23:59Step 2 Soupedup Screenreaders
- 25:07Step 3 Robbie the Reader
- 25:45Conversational Interface
- 26:11Disappointment
- 26:55Web API
- 27:24The Internet
- 28:50Screen Readers vs VoiceHTML
- 29:33Conflict of Interest
- 30:20Final Thoughts
- 31:38Contact Information
related talks
46mThe Sound of Silence: Web APIs for Accessibility for Visual and Hearing Impairments
29mAccessibility vs latest Web APIs. Can’t we just get along?
58mAdding Speech to your Web App_UIArchitecture
31mBringing the Web Back to the Blind
49mDesigning voice interfaces
51mDesigning the Conversation
48mConversational ULS: Talking to Siri, Alexa, and Your Web Browser
38mIntegrating AI and Browser API Voice Recognition on Web App