Whisper is a general-purpose speech recognition model. It is trained on a large dataset of diverse audio and is also a multitasking model that can perform multilingual speech recognition, speech translation, and language identification.

FMAR Whisper is an unmodified copy of whisper.unity. It has been packaged in order to create dependencies, as it does not use the package system or assembly definitions. We have added the necessary assembly definitions for these scripts and made them compatible with FMAR.

More information about whisper.unity can be found here: https://github.com/Macoron/whisper.unity whisper.unity uses the MIT License and is therefore freely modifiable and redistributable.