Controllable Game Level Generation: Assessing the Effect of Negative Examples in GAN Models
This work addresses the challenge of generating specific game levels for developers, but it is incremental as it builds on existing controllable GAN methods.
The paper tackled the problem of generating controllable game levels using GAN models, specifically comparing CGAN and Rumi-GAN with and without negative examples to enforce constraints like playability and controllability, finding that negative examples help avoid undesirable outputs but with varying strengths and weaknesses.
Generative Adversarial Networks (GANs) are unsupervised models designed to learn and replicate a target distribution. The vanilla versions of these models can be extended to more controllable models. Conditional Generative Adversarial Networks (CGANs) extend vanilla GANs by conditioning both the generator and discriminator on some additional information (labels). Controllable models based on complementary learning, such as Rumi-GAN, have been introduced. Rumi-GANs leverage negative examples to enhance the generator's ability to learn positive examples. We evaluate the performance of two controllable GAN variants, CGAN and Rumi-GAN, in generating game levels targeting specific constraints of interest: playability and controllability. This evaluation is conducted under two scenarios: with and without the inclusion of negative examples. The goal is to determine whether incorporating negative examples helps the GAN models avoid generating undesirable outputs. Our findings highlight the strengths and weaknesses of each method in enforcing the generation of specific conditions when generating outputs based on given positive and negative examples.