Drupal & AI Adventures: Drupal as a Dictaphone

Hello Drupal fanatics! Welcome to “Drupal & AI Adventures. In this series, we explore the fantastic world of Drupal modules powered by artificial intelligence. Our mission? We want to inspire and enlighten YOU. We'll dive in together! This time, we're talking about Speech to text. 

The following modules allow Drupal to be used as a dictaphone. 

• You speak and the text is entered into a text field.
  • https://www.drupal.org/project/cke_5_speech_to_type
  • This module is not widely used yet, good interface, clean code.
• Uses the following browser api for speech to text:(window.speechRecognition || window.webkitSpeechRecognition). 
  • https://www.drupal.org/project/speech_to_text
  • Similar module to the above, uses the same browser API. Code is slightly more concise as above.

The following modules allow you to turn audio into text. You load audio and text comes out.

• https://www.drupal.org/project/augmentor_google_cloud_speech_to_text
  • This module fits into the Augmentor module ecosystem, a separate blog post on this will follow.
• https://www.drupal.org/project/google_stt 
  • This module is looking for new maintainers and depends on google cloud to do speech to text. 
• https://www.drupal.org/project/openai 
  • The Openai module also includes a submodule “Openai speech to text” that allows you to convert audio files to text. 
• https://www.drupal.org/project/ai_interpolator_deepgram
  • In the AI interpolator module ECO system there is also an integration with Deepgram. This integration has following cool features:
    • Diarization - It hears which person is speaking.
    • Timecode - Puts a timecode to each sentence.
    • Text Split - Can split the text into separate text fields if desired. 

Of all the modules, this is the most powerful I think.