MightyTranscription

📦 What is MightyTranscription?

MightyTranscription is a powerful Unreal Engine plugin that converts audio input into accurate text using OpenAI's Whisper library. Perfect for in-game transcription, subtitles, voice commands, and more.

🚀 How to Install

1. Download the plugin from the Fab store.
2. Enable MightyTranscription in the Plugin Manager.
3. Restart the Unreal Engine Editor.

🔧 How to Use

After enabling the plugin:

• Open a blueprint where you want to convert the audio to text
• Right click the blueprint editor and search for 'Download Model'. Now create a Mighty Transcription > 'Download Model' async blueprint node
• Link the node to the 'BeginPlay' event in your blueprint or anywhere else which will be executed
• Set the 'ModelName' parameter to one of the models at https://huggingface.co/ggerganov/whisper.cpp or use the default 'base.en' model
• Right click the blueprint editor and search for 'Transcribe Input File'. Create a 'Transcribe Input File' blueprint node
• Create an FString variable e.g. MyTranscribedText to store the transcribed text
• Create node to set the value of the FString variable e.g. Set MyTranscribedText
• Link the 'OnSuccess' output to the input execution pin of the 'Set MyTranscribedText' node.
• Drag from the 'Params' input on the 'Transcribe Input File' node and place a 'Make Transcription Params' node
• Link the 'ModelPath' output pin on the 'Download Model' async node to the 'ModelPath' property in the Transcription Params struct
• Now set the 'Input File Path' property in the Transcription Params struct to the path of your audio file e.g C:\myaudio.wav
• Executing this blueprint code should now convert the audio file to an FString and store the output in 'MyTranscribedText' variable

Your final blueprint should look like:

💬 Support

If you have any questions, suggestions or issues, feel free to raise a github issue at https://github.com/mustardpower/mighty-transcription-docs/issues