Memoria: Keeping Family Voices Alive with Local Whisper + Gemma
说明与用途
Memoria 是一个本地口述历史档案工具,使用 Whisper 和 Gemma 对家庭语音备忘录进行转录和结构化摘要,所有数据均保存在本地。
问题
Family voice memos contain sensitive personal data that should not be sent to cloud services for transcription and summarization.
结果
Memoria achieves fully local, privacy-preserving transcription and structured summarization of family voice memos, producing JSON and Markdown archives within ~30 s on a local CPU.
工作流
- Upload or record a family voice memo
- Run Whisper small (CPU int8) to generate transcript with timestamps
- Feed transcript to Gemma 3 4B (llama.cpp) to produce structured fields (title, summary, people, places, time references, themes, quotes, follow-up questions)
- Apply deterministic quote verification to ensure quotes exist in transcript
- Export results as JSON and Markdown
- Optionally view via the FastAPI web interface
证据
Memoria runs both speech recognition and language understanding locally. The recording and transcript never leave the machine for inference.
The bundled 58‑second story is synthetic and fictional, narrated with TTS specifically so that no private family recording is exposed.
I ran the project rather than stopping at a screenshot or a plausible demo response: Unit tests 6/6 passed, demo audio duration 57.79 seconds, full CLI run 33.36 seconds.
- 记录日期: 2026-10-27T09:11:52.202Z