Hi! I have a small idea that might be interesting for the project.
It would be great if the AI character could see what is happening through a camera in real time, understand the user's actions, and react accordingly — for example, adjusting its dialogue or actions based on what it sees.
Something similar has already been implemented in another project. I'll leave a link below as an example/reference: https://github.com/blucrew/VisualStimEdger