cedar inlet
I'm not sure how Neuro's facial expression engine works, but I had an idea that Vedal, or someone in the swarm, or hired help can create the thing that collects data from vbridger, ARKit, OBS, toggles and audio for training Neuro.
The thing being a script that runs in the background or a plugin that can be implemented or something else.
We'll just have to do some mental gymnastics and pretend that it's not just scraping data from fellow vtubers and Neuro is just watching their streams 
Consent from those vtubers goes without saying. (oh no, Vedal will have to talk to someone 
)
Data quality? Outsource cleanup to the Swarm or embrace the scuff.
Actually, if the thing will be able to stream the data, and Vedal comes up with a workflow we might have Neuro in chat sometimes. Also, collabs might be a little bit better if she has more grasp on their emotional state real-time.
lunar schooner
