8,464
edits
Restless Souls/Technology: Difference between revisions
m
→Neural link
Paradox-01 (talk | contribs) m (research on speech synthesis made some steps forwards but still has a way to go) |
Paradox-01 (talk | contribs) m (→Neural link) |
||
| Line 1,630: | Line 1,630: | ||
* collecting visual and auditory patterns | * collecting visual and auditory patterns | ||
* collecting weak patterns of muscle memory for speech ("inner voice") | * collecting weak patterns of muscle memory for speech ("inner voice") | ||
** [https://www.youtube.com/watch?v=3pv0vT82Cys current RL research is still busy trying to reconstruct speech from patterns originating more ''higher up'' in the brain] | ** [https://www.youtube.com/watch?v=3pv0vT82Cys current RL research is still busy trying to reconstruct speech from patterns originating more ''higher up'' in the brain] [https://www.nature.com/articles/s41586-019-1119-1 (BBC)], [https://www.nature.com/articles/s41586-019-1119-1 (nature)] | ||
** Note this part: "Even when participants mouthed sentences but did not speak them out loud, the model produced an intelligible synthesized sentence." | |||
* collecting weak patterns of muscle memory for gestures and movements ("imagined moves") | * collecting weak patterns of muscle memory for gestures and movements ("imagined moves") | ||
* collecting any other neural pattern (as long as it isn't too complex for analysis such as pain) | * collecting any other neural pattern (as long as it isn't too complex for analysis such as pain) | ||