Generic Guard AI in Stealth Game with Composite Potential Fields

2025-08-25

AI Summary

In stealth games, designing guard patrol behaviors that simultaneously achieve high coverage efficiency, responsive pursuit, and behavioral naturalness remains challenging. Method: This paper proposes a training-free, fully interpretable general-purpose guard AI framework. It introduces three novel spatial fields—information field, confidence field, and connectivity field—integrated via composite potential fields to fuse global map knowledge with local perception. Leveraging kernel filtering and parameterized design, the framework incorporates interference-response and environment-interaction modules, enabling zero-shot dynamic behavior tuning. Contribution/Results: Evaluated across five diverse map types and multiple adversarial strategies, the framework achieves a 23.6% improvement in target capture rate and a 37% increase in expert-rated behavioral naturalness. It supports heterogeneous map abstractions and guard modalities, facilitating rapid prototyping and deployment in varied stealth game scenarios.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Guard patrol behavior is central to the immersion and strategic depth of stealth games, while most existing systems rely on hand-crafted routes or specialized logic that struggle to balance coverage efficiency and responsive pursuit with believable naturalness. We propose a generic, fully explainable, training-free framework that integrates global knowledge and local information via Composite Potential Fields, combining three interpretable maps-Information, Confidence, and Connectivity-into a single kernel-filtered decision criterion. Our parametric, designer-driven approach requires only a handful of decay and weight parameters-no retraining-to smoothly adapt across both occupancy-grid and NavMesh-partition abstractions. We evaluate on five representative game maps, two player-control policies, and five guard modes, confirming that our method outperforms classical baseline methods in both capture efficiency and patrol naturalness. Finally, we show how common stealth mechanics-distractions and environmental elements-integrate naturally into our framework as sub modules, enabling rapid prototyping of rich, dynamic, and responsive guard behaviors.

Problem

• Designing believable guard patrol behavior in stealth games
• Balancing coverage efficiency and responsive pursuit naturally
• Creating a generic training-free framework for guard AI

Innovation

• Composite Potential Fields framework
• Integrates global and local information
• Parametric designer-driven training-free approach