V1.1.0

In this new version of LVATT, many features have been introduced, most of which address the previous drawbacks of the projects. They are:

• Custom and multiple file input. Which allows for you to process x amount of files efficiently.
• Whisper model input, which allows for either custom model with path or preset model, which will get automatically downloaded if not already.

Some planned features:

• Parallel file processing.
• Multiple types of demodulation

Full Changelog: V1.0.0...V1.1.0

V1.0.0

First fully working version.

Features:

• Takes in an IQ file (FM modulated) and outputs a WAV file containing the demodulated audio
• Takes the demodulated audio straight from memory (none of that reading for disk shiiiii), transcribes and translates into text

Current Limitations:

• The whole project is mostly hard coded (file name, only 1 file input, sample rate, language etc), We plan on turning this into a very flexible and intuitive tool.
• Due to there not being too many resources and DSP being a very big field, and so the project isn't as flexible as it could be (AM demodulation, etc)