A Bayesian Model for Plan Recognition in RTS Games applied to StarCraft

11/16/2011
by   Gabriel Synnaeve, et al.
0

The task of keyhole (unobtrusive) plan recognition is central to adaptive game AI. "Tech trees" or "build trees" are the core of real-time strategy (RTS) game strategic (long term) planning. This paper presents a generic and simple Bayesian model for RTS build tree prediction from noisy observations, which parameters are learned from replays (game logs). This unsupervised machine learning approach involves minimal work for the game developers as it leverage players' data (com- mon in RTS). We applied it to StarCraft1 and showed that it yields high quality and robust predictions, that can feed an adaptive AI.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
07/01/2022

The NPC AI of The Last of Us: A case study

The Last of Us is a game focused on stealth, companionship and strategy....
research
05/14/2023

Introducing Tales of Tribute AI Competition

This paper presents a new AI challenge, the Tales of Tribute AI Competit...
research
03/25/2019

Winning Isn't Everything: Training Human-Like Agents for Playtesting and Game AI

Recently, there have been several high-profile achievements of agents le...
research
03/04/2021

The Dota 2 Bot Competition

Multiplayer Online Battle Area (MOBA) games are a recent huge success bo...
research
02/25/2019

Similarity Measures based on Local Game Trees

We study strategic similarity of game positions in two-player extensive ...
research
09/18/2020

TotalBotWar: A New Pseudo Real-time Multi-action Game Challenge and Competition for AI

This paper presents TotalBotWar, a new pseudo real-time multi-action cha...
research
12/06/2021

On the complexity of Dark Chinese Chess

This paper provides a complexity analysis for the game of dark Chinese c...

Please sign up or login with your details

Forgot password? Click here to reset