@inproceedings{612402292b994c28824e8775b64a947c,
title = "Path of Destruction: Learning an Iterative Level Generator Using a Small Dataset",
abstract = "We propose a new procedural content generation method which learns iterative level generators from a dataset of existing levels. The Path of Destruction method, as we call it, views level generation as repair; levels are created by iteratively repairing from a random starting level. The first step is to generate an artificial dataset from the original set of levels by introducing many different sequences of mutations to existing levels. In the generated dataset, features are observations of destroyed levels and targets are the specific actions that repair the mutated tile in the middle of the observations. Using this dataset, a convolutional network is trained to map from observations to their respective appropriate repair actions. The trained network is then used to iteratively produce levels from random starting maps. We demonstrate this method by applying it to generate unique and playable tile-based levels for several 2D games (Zelda, Danger Dave, and Sokoban) and vary key hyperparameters.",
keywords = "data augmentation, games, level generation, neural networks, procedural content generation, supervised learning",
author = "Matthew Siper and Ahmed Khalifa and Julian Togelius",
note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022 ; Conference date: 04-12-2022 Through 07-12-2022",
year = "2022",
doi = "10.1109/SSCI51031.2022.10022073",
language = "English (US)",
series = "Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "337--343",
editor = "Hisao Ishibuchi and Chee-Keong Kwoh and Ah-Hwee Tan and Dipti Srinivasan and Chunyan Miao and Anupam Trivedi and Keeley Crockett",
booktitle = "Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022",
}