Path of Destruction: Learning an Iterative Level Generator Using a Small Dataset
This addresses the challenge of creating diverse game levels with limited data, though it is incremental as it builds on existing repair-based generation ideas.
The authors tackled the problem of procedural content generation by proposing the Path of Destruction method, which learns iterative level generators from small datasets, resulting in the generation of unique and playable tile-based levels for games like Zelda, Danger Dave, and Sokoban.
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.