Three-dimensional attention Transformer for state evaluation in real-time strategy games

TL;DR

This paper introduces a tri-dimensional Transformer architecture for real-time strategy game situation assessment, effectively modeling spatial, temporal, and feature information to improve accuracy and efficiency.

Contribution

It proposes a novel tri-dimensional Space-Time-Feature Transformer that outperforms existing models in RTS game state evaluation with fewer parameters.

Findings

Achieves 58.7% accuracy early game, surpassing Timesformer.

Reaches 97.6% mid-game accuracy with low variation.

Uses fewer parameters than baseline models.

Abstract

Situation assessment in Real-Time Strategy (RTS) games is crucial for understanding decision-making in complex adversarial environments. However, existing methods remain limited in processing multi-dimensional feature information and temporal dependencies. Here we propose a tri-dimensional Space-Time-Feature Transformer (TSTF Transformer) architecture, which efficiently models battlefield situations through three independent but cascaded modules: spatial attention, temporal attention, and feature attention. On a dataset comprising 3,150 adversarial experiments, the 8-layer TSTF Transformer demonstrates superior performance: achieving 58.7% accuracy in the early game (~4% progress), significantly outperforming the conventional Timesformer's 41.8%; reaching 97.6% accuracy in the mid-game (~40% progress) while maintaining low performance variation (standard deviation 0.114). Meanwhile, this architecture requires fewer parameters (4.75M) compared to the baseline model (5.54M). Our study not only provides new insights into situation assessment in RTS games but also presents an innovative paradigm for Transformer-based multi-dimensional temporal modeling.