Multi-agent reinforcement learning framework for training NPCs in browser-based 3D voxel hide-and-seek using PPO and WebSocket communication between Ray RLlib and THREE.js
webgl threejs machine-learning reinforcement-learning computer-vision deep-learning procedural-generation websocket pytorch voxel-engine game-ai multi-agent-systems hide-and-seek proximal-policy-optimization ppo 3d-games self-play ray-rllib npc-behavior browser-based-ml
-
Updated
Dec 3, 2025 - JavaScript