Provides live closed captioning and speaker identification for video rooms with a customizable floating overlay.
The ai-captions skill enables real-time transcription and subtitle overlays for video conferencing applications. By integrating Deepgram transcription with LiveKit data channels, it delivers low-latency captions directly to participants. The skill features a professional UI suite including a floating overlay, speaker labeling, and a comprehensive settings panel that allows users to adjust font size, positioning, and background opacity for maximum accessibility and readability.
主要功能
01Reusable React hooks for managing caption state and timing
02Real-time transcription via Deepgram and LiveKit data channels
033 GitHub stars
04Intelligent caption fading and text truncation
05Customizable UI settings for font size, position, and opacity
06Automatic speaker identification with formatted labels
使用场景
01Enhancing accessibility for video conferencing and virtual meetings
02Improving user experience in multi-speaker calls with visual speaker cues
03Providing live subtitles for webinars and online educational platforms
