video-transcribe

To transcribe a YouTube video, the skill extracts audio via yt-dlp and processes it to generate timestamped transcripts. This produces accurate text notes and visual keyframes for summarizing educational lectures.

Yes, the skill extracts visual keyframes and generates scene descriptions from video content. This visual analysis provides textual descriptions of frames, enabling you to review media content without watching the entire source video.

Yes, yt-dlp and ffmpeg are required dependencies for downloading and processing video media. You also need the Groq Whisper API for online transcription, or whisper-cpp installed locally if you require offline processing.

The best way to summarize lengthy lecture videos is by transcribing spoken audio and extracting visual keyframes. This workflow transforms long educational media into concise, summarized notes with embedded screenshots for quick reference.

Yes, video transcription works offline using whisper-cpp as an alternative to the Groq Whisper API. This allows you to process audio and generate timestamped transcripts locally without requiring an active internet connection for media analysis.

You create an Obsidian note by processing the video to extract timestamped transcripts and key frames. The skill formats the transcribed spoken words and visual scene descriptions into a detailed markdown note for educational media review.