you will need to tape some radio shows or copy and paste some podcast content into a folder for the basic stems to train the ai model on.
In fact this is a dual model ai.
We will also build up a small library of text, so we can run the text to speech AND the speech to text bi-directionally .
Finally, we will build the voice commands for a speech to text, to command processor.
Although this will only be able to achieve a small handful of voice commands but the skills learned here will contribute to far more complex AI projects further down the line.
However that’s still not so bad for our first AI model lets proceed to the projet.
If you want to record the podcast or lecture yourself you can use an x-box controller, plug in a video game headset and plug this into the computer. You will need to use OBS streamer and give windows permissions to use the microphone somewhere within windows settings.
One of the more common models from Nvidia ai will allow you to work speech to text against the Nvidia AI model but you will need to pay about a dollar an hour to use it from the Nvidia cloud services.