aTrainTranscription/aTrain

A GUI tool for offline transcription of speech recordings, including speaker diarization, utilizing state-of-the-art machine learning models.

What it solves

aTrain provides a private, offline way to transcribe speech recordings into text. It specifically addresses the need for researchers to handle sensitive interview data while complying with privacy regulations like GDPR, as no data is uploaded to the internet.

How it works

The tool leverages the faster-whisper implementation of OpenAI's Whisper model for high-quality transcription and uses pyannote.audio for speaker detection (diarization). It can run locally on a device's CPU or an NVIDIA GPU via CUDA to accelerate processing. It is available as a desktop application (Windows/Linux) and as a CLI tool for headless automation.

Who it’s for

It is primarily designed for researchers conducting qualitative analysis who need to transcribe interviews and import the results into analysis software like MAXQDA, ATLAS.ti, or nVivo.

Highlights

  • Privacy-First: Processes all recordings completely offline on the local device.
  • Speaker Detection: Identifies different speakers within a recording to attribute text segments.
  • Broad Language Support: Supports transcription for 99 different languages.
  • Analysis Tool Integration: Generates output files compatible with popular qualitative analysis software.
  • Hardware Acceleration: Supports NVIDIA GPUs to significantly reduce transcription time.

Related

  • Project
    jasoncheng7115/jt-live-whisper

    jt‑live‑whisper is an offline‑first AI toolbox for real‑time transcription, translation, speaker diarisation and meeting summarisation. It captures system audio on macOS, Windows or Linux, runs Whisper‑based or other ASR models locally, translates with a local LLM (or offline NLLB/Argos), and can produce subtitles, transcripts and AI‑generated meeting minutes—all without any cloud API.

  • Project
    OpenWhispr/openwhispr

    A privacy-first, open-source voice-to-text application providing local AI dictation, meeting transcription with speaker diarization, and AI-powered notes.

  • Project
    mybigday/whisper.rn

    whisper.rn is a React Native library that brings OpenAI’s Whisper (and NVIDIA Parakeet) speech‑to‑text models to iOS and Android. It ships pre‑built native binaries, supports Core ML, GPU, and Snapdragon Hexagon NPU acceleration, includes a built‑in VAD, and provides a realtime transcription API, all usable offline from JavaScript.

  • Project
    openai/whisper

    Whisper is OpenAI’s open‑source, transformer‑based speech‑to‑text system that handles transcription, language identification, and translation across many languages. It ships as a pip‑installable package with six model sizes (tiny → large, plus a fast “turbo” variant). The CLI (`whisper …`) and a simple Python API (`whisper.load_model(...).transcribe(...)`) let you turn audio files into text with just a few commands. The project runs on Python 3.8‑3.11, PyTorch, and requires ffmpeg (and optionally Rust for the tokenizer). All code and model weights are MIT‑licensed.

  • Project
    cjpais/Handy

    A free, open-source, and offline speech-to-text application that transcribes voice to text directly into any active text field on your computer.