TY - JOUR

T1 - Wor(l)d-GAN

T2 - Towards Natural Language Based PCG in Minecraft

AU - Awiszus, Maren

AU - Schubert, Frederik

AU - Rosenhahn, Bodo

N1 - This work was supported by the Federal Ministry of Education and Research (BMBF), Germany under the project LeibnizKILabor (grant no. 01DD20003) and the Deutsche Forschungsgemeinschaft (DFG) under Germany’s Excellence Strategy within the Cluster of Excellence PhoenixD (EXC 2122).

PY - 2023/6

Y1 - 2023/6

N2 - This work presents Wor(l)d-GAN, a method to perform data-driven Procedural Content Generation via Machine Learning in Minecraft from a single example. Based on a 3D Generative Adversarial Network (GAN) architecture, we are able to create arbitrarily sized world snippets from a given sample. Our method applies dense representations used in Natural Language Processing (NLP) in two ways. First, we propose block2vec representations based on word2vec. Secondly, we use the pretrained large language model BERT to generate representations directly from the token names. These representations make Wor(l)d-GAN independent from the number of different blocks, which can vary a lot in Minecraft, and enable the generation of larger levels. We evaluate our approach on creations from the community as well as structures generated with the Minecraft World Generator under several metrics. Wor(l)d-GAN enables its users to generate Minecraft worlds based on parts of their creations.

AB - This work presents Wor(l)d-GAN, a method to perform data-driven Procedural Content Generation via Machine Learning in Minecraft from a single example. Based on a 3D Generative Adversarial Network (GAN) architecture, we are able to create arbitrarily sized world snippets from a given sample. Our method applies dense representations used in Natural Language Processing (NLP) in two ways. First, we propose block2vec representations based on word2vec. Secondly, we use the pretrained large language model BERT to generate representations directly from the token names. These representations make Wor(l)d-GAN independent from the number of different blocks, which can vary a lot in Minecraft, and enable the generation of larger levels. We evaluate our approach on creations from the community as well as structures generated with the Minecraft World Generator under several metrics. Wor(l)d-GAN enables its users to generate Minecraft worlds based on parts of their creations.

KW - BERT

KW - GAN

KW - Generation

KW - Level

KW - Minecraft

KW - NLP

KW - PCG

KW - Representation

KW - Scales

KW - SinGAN

KW - Single Example

KW - Style

KW - generation

KW - natural language processing (NLP)

KW - generative adversarial network (GAN)

KW - level

KW - single example

KW - representation

KW - scales

KW - procedural content generation (PCG)

KW - style

UR - http://www.scopus.com/inward/record.url?scp=85125353461&partnerID=8YFLogxK

U2 - 10.1109/tg.2022.3153206

DO - 10.1109/tg.2022.3153206

M3 - Article

AN - SCOPUS:85125353461

VL - 15

SP - 182

EP - 192

JO - IEEE Transactions on Games

JF - IEEE Transactions on Games

SN - 2475-1502

IS - 2

ER -