The motivation for building a custom karaoke game was to expand the song selection for a favorite band and add a scoring mechanic.
The Web Speech API is a native browser feature split into speech recognition (speech-to-text) and speech synthesis (text-to-speech) for voice input and control.
The API's real-world usability is hampered by inconsistent browser support, reliance on vendor-specific cloud services, and privacy concerns.
A live demonstration shows how the API can be used for simple tasks like reading selected text aloud or enabling basic voice commands on a webpage.
The game's implementation involves initializing speech recognition, handling its tendency to stop by restarting it, and comparing the transcribed text against the song's lyrics.
The completed karaoke game is tested by speaking, singing, and intentionally using incorrect words to show how it scores the performance.
While the native API is not yet production-ready for commercial use, the community has built useful tools and projects like Web Captioner and Mozilla's Common Voice.
When designing for voice, it's crucial to use concise content, consider conversational flow, and plan for error recovery and cognitive load.
Building "useless" or "silly" side projects is a powerful way to combat burnout, learn new skills, and rediscover the fun in software development.
Sensory-Minds GmbH
Offenbach am Main, Germany
28:09Ana Rodrigues
41:03Nolan Royalty
23:09Tobias Münch
32:54Rowdy Rabouw
26:28Anna Oliveira
32:54
30:59Lee Boonstra
1:13:35Chris Heilmann & Daniel Cranney & Rowdy Rabouw
Jobs that call for the skills explored in this talk.
Synthesia Limited
Charing Cross, United Kingdom
