Addressing the Fundamental Tension of PCGML with Discriminative Learning

TL;DR

This paper introduces a discriminative learning approach for PCGML that emphasizes validity over distribution, enabling more efficient and interactive content generation with user-guided positive and negative examples.

Contribution

It proposes using discriminative models trained on positive and negative examples to improve PCGML control and interaction, modifying WaveFunctionCollapse for this purpose.

Findings

Discriminative models effectively capture design validity.

Enhanced control through positive and negative example critique.

Bridges PCGML with mixed-initiative design tools.

Abstract

Procedural content generation via machine learning (PCGML) is typically framed as the task of fitting a generative model to full-scale examples of a desired content distribution. This approach presents a fundamental tension: the more design effort expended to produce detailed training examples for shaping a generator, the lower the return on investment from applying PCGML in the first place. In response, we propose the use of discriminative models (which capture the validity of a design rather the distribution of the content) trained on positive and negative examples. Through a modest modification of WaveFunctionCollapse, a commercially-adopted PCG approach that we characterize as using elementary machine learning, we demonstrate a new mode of control for learning-based generators. We demonstrate how an artist might craft a focused set of additional positive and negative examples by critique of the generator's previous outputs. This interaction mode bridges PCGML with mixed-initiative design assistance tools by working with a machine to define a space of valid designs rather than just one new design.