Flatsonar

aTrain

by Armin Haberl, Jürgen Fleiß

Offline transcription of speech recordings

Broad permissions Verified creator AGPL-3.0-or-later v1.4.2
Screenshot 1 of aTrain Screenshot 2 of aTrain Screenshot 3 of aTrain Screenshot 4 of aTrain Screenshot 5 of aTrain

About

aTrain is a tool for automatically transcribing speech recordings utilizing state-of-the-art machine learning models without uploading any data. It was developed by researchers at the Business Analytics and Data Science-Center at the University of Graz and tested by researchers from the Know-Center Graz.

Features include:

  • Automatic speech-to-text transcription
  • Speaker diarization
  • GPU acceleration with CUDA support
  • Export transcriptions in various formats, including MAXQDA, ATLAS.ti and nVivo compatible output,
  • Completely offline - all processing done locally

Permissions

1 permission weaken the sandbox. Flatsonar asks before installing.

  • X11 display: X11 lets apps read input and windows of other apps
    --socket=x11
  • network
    --share=network
  • inter-process communication (needed for X11)
    --share=ipc
  • Wayland display
    --socket=wayland
  • X11 display, only when Wayland is unavailable
    --socket=fallback-x11
  • GPU acceleration
    --device=dri
  • access to your documents folder
    --filesystem=xdg-documents

Who publishes this

The creator demonstrably controls this app id: Flathub verification, the hosting account owns the namespace, or a well-known file on their domain.

  • verified on Flathub: the developer proved they control this app id
    publisher:flathub

Maintenance

Recent activity on the upstream repository, or built and reviewed by Flathub.

No staleness signals: recent activity, or built and reviewed by Flathub.