Wearable Live Captions
summary
A wearable LED mask displays live captions using Azure Cognitive Services speech‑to‑text streamed from a browser microphone, packs commands into a tight binary format, and publishes them via the Ably MQTT broker to a tiny microcontroller that renders scrolling text on a 28×8 neopixel array. Font bitmaps and an index live on the device to avoid heavy string parsing, scrolling runs by decrementing an X offset, the build fits in a mask, costs under ~$15 and demonstrably works in real interactions.
AI-written summary · about
chapters (8)
- 0:00Introduction and setup
- 4:02Joe’s mother’s hearing loss
- 10:23Building the wearable display
- 16:48Efficient data transmission
- 21:20Text rendering on hardware
- 28:45Scrolling text implementation
- 33:15Demo and user reaction
- 36:44Conclusion and Q&A
related talks
38mWearable Live Captions
49mMaking IoT: An Intro for Web Devs
51mKeyboards? Where we’re going, we don’t need keyboards
28mCode you can hold: Making your first IoT wearable
16mAlterWear: Battery-Free Wearable Displays for Opportunistic Interactions
20mMake an Amazing World pt.2
25mHaute Codeture
19mHow Captions Enables Anyone To Tell Their Story With Video