Collective Evolution Learning Model for Vision-Based Collective Motion with Collision Avoidance
David L. Krongauz, Teddy Lazebnik
Q-learning policy trained on genetic-algorithm targets lets simulated locust-shaped agents with only visual input flock (CM score 0.7) and avoid collisions (CA 0.74 vs 0.17 for a flocking-only policy).
PLoS ONE · May 10, 2023Peer-reviewedAnimal Behavior & Cognition✦ In-depth digest