New release for Transcribable now available for open testing show casing our offline speech recogniser – Android OS 1.12.04

The Transcribable 1.12.x branch introoduces several features. The most significant of which is our new offline capable speech recogniser; it uses a combination of Whisper.cpp and Silero VAD to achieve this.

Changelog

Android OS app (369011204)

• Offline capable speech recogniser (experimental; feedback welcome, note this prototype may crash or be unreliable although we have tested it there many be scenarios we haven’t covered.):
  • Uses Whisper.cpp for speech recognition; provides support for downloading several different models (ggml-*). These options are currently limited to a selected few models but it will be expanded; if there is a particular Whisper model of interest let us know.
  • Silero VAD is used for voice activity detection.
  • Models for use with: Silero + Whisper are not bundled with the app; and require an additional download within the application.
    • This is primarily to avoid inflating the applications size with assets (models in this instance) that the user may not use. As Transcribable is intended to be used with many speech recogniser applications.
  • Our speech recogniser has a dedicated preference activity allowing:
    • Model downloads and changes.
    • Manage and remove downloaded models.
  • Features a continuous transcription mode; that processes results when speech activity pauses.
    • Side note this works especially well with when using a Samsung S-Pen (with remote S-pen functionality) can use the pen’s action button to start and stop the recording.
• Tasker plugin support
  • Send text to the current file or a separate file (created if it doesn’t already exist).
  • Retrieve text from the current active file.
• UI improvements:
  • Floating action button now cradles to the side of the toolbar.
  • Speech recogniser listening state adds additional animation to background of editor.
    • Can be disable through a preference.
    • Has extra animations stages for our new speech recogniser.
  • Various improvements for SDK 36 target involving insets.

Learn more about Transcribable.

This update is currently only available on the open (and internal) testing track; the link below will take you to open testing sign-up page.

Before this is released to the public track; here are a few items we plan to expand on:

• Expand models available under model selection.
  • Allow models to be exported/imported from another storage location to avoid re-downloading.
• Add speech recognition activity for use outside Transcribable (system voice input).
• Add support to allow audio files to be converted (within the app and via sharing intent from external apps).
• Optimise and improve the offline recognition utility further.

We hope you enjoy the latest update!