Boosting Inclusivity: Closed Captioning & Translations in WebRTC
summary
Demo shows a working backend that centralizes audio from a WebRTC media server, transcodes webm Opus → 16 kHz 16‑bit mono PCM WAV with FFmpeg, and streams it to a cloud ASR that provides realtime transcription and translation. Implementation uses the JavaScript SDK and the Azure Speech conversation translator class when possible, runs continuous mode (requires restarts to avoid timeouts), and hits a cross‑region timeout bug; recordings with subtitles remain on the to‑do list.
AI-written summary · about
chapters (8)
- 0:00Introduction to speaker
- 1:09What is speech recognition
- 3:20Benefits of transcription
- 5:46Choosing a speech service
- 9:03WebRTC app setup
- 12:59Backend integration details
- 17:38Results and demo
- 18:49Q&A session
related talks
45mAccessible Audio & Video
31mShare information by removing language barriers with Amazon Translate
54mLive Captioning: Make your next event accessible for everyone
09mWeb apps that talk
29mUsing Amazon Transcribe to make content searchable and accessible
29mAccessibility vs latest Web APIs. Can’t we just get along?
48mProgramming for Accessibility
46mThe Sound of Silence: Web APIs for Accessibility for Visual and Hearing Impairments