Learning Dark Souls Combat Through Pixel Input With Neuroevolution

TL;DR

This paper demonstrates that neuroevolution can be used to train agents to play Dark Souls directly from pixel inputs, using a novel API for real-time game data extraction, achieving notable success in complex combat scenarios.

Contribution

The paper introduces a neuroevolution approach with a new game API to enable vision-based learning in complex, high-dimensional game environments like Dark Souls.

Findings

Agents achieved up to 35% success rate in defeating the boss

Neuroevolution effectively learns combat strategies from raw pixel data

The approach bypasses the need for explicit game-state information

Abstract

This paper investigates the application of Neuroevolution of Augmenting Topologies (NEAT) to automate gameplay in Dark Souls, a notoriously challenging action role-playing game characterized by complex combat mechanics, dynamic environments, and high-dimensional visual inputs. Unlike traditional reinforcement learning or game playing approaches, our method evolves neural networks directly from raw pixel data, circumventing the need for explicit game-state information. To facilitate this approach, we introduce the Dark Souls API (DSAPI), a novel Python framework leveraging real-time computer vision techniques for extracting critical game metrics, including player and enemy health states. Using NEAT, agents evolve effective combat strategies for defeating the Asylum Demon, the game's initial boss, without predefined behaviors or domain-specific heuristics. Experimental results demonstrate that evolved agents achieve up to a 35% success rate, indicating the viability of neuroevolution in addressing complex, visually intricate gameplay scenarios. This work represents an interesting application of vision-based neuroevolution, highlighting its potential use in a wide range of challenging game environments lacking direct API support or well-defined state representations.