Anyone have experience with Utility AI using Influence Maps and LOS? I have a use case where designers want AI to pick the best spot to attack a target from. I could render a basic LOS mask from the POV of the target and then use that as a mask, but expensive for multi agents...
Could do something akin to 2d visibility (see: https://www.redblobgames.com/articles/visibility/ …) but the map instead represents how *many* agents belonging to a certain team are visible from a given spot. If using GPU acceleration, stencil op add could be very effective here.
-
-
Describing it in tweets is hard to nail down. Going to attempt to mock up an example to show what I mean. Will be a moment.
-
This is the idea. Red are blockers, blue are agents of a given team. Each team gets 1 map. Redraw the influence map every frame (or nth frame). Let's AI maximize (big AoE) or minimize (seek cover) LOS. Actual targeting selection would be separate operation.pic.twitter.com/OIHIJTBx4q
- Näytä vastaukset
Uusi keskustelu -
Lataaminen näyttää kestävän hetken.
Twitter saattaa olla ruuhkautunut tai ongelma on muuten hetkellinen. Yritä uudelleen tai käy Twitterin tilasivulla saadaksesi lisätietoja.