ParamExplorer: A framework for exploring parameters in generative art

2025-12-18

AI Summary

In generative art, the high-dimensional, sparse, and fragmented parameter space impedes efficient discovery of aesthetically high-quality outputs, rendering traditional manual trial-and-error approaches inefficient. To address this, we propose the first interactive parameter exploration framework integrating reinforcement learning principles, enabling both human-in-the-loop collaboration and fully automated, feedback-driven search, with native p5.js integration. The framework features a modular, plug-and-play multi-strategy agent architecture that unifies stochastic sampling, user preference modeling, and Bayesian optimization. Its web-based real-time visualization infrastructure significantly accelerates exploration. Evaluated across multiple generative algorithms, user studies demonstrate a >40% reduction in exploration time and a threefold increase in the discovery rate of novel, high-quality configurations—marking a paradigm shift from empirical trial-and-error to goal-directed, feedback-optimized exploration.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Natural Language Processing: Generation

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Generative art systems often involve high-dimensional and complex parameter spaces in which aesthetically compelling outputs occupy only small, fragmented regions. Because of this combinatorial explosion, artists typically rely on extensive manual trial-and-error, leaving many potentially interesting configurations undiscovered. In this work we make two contributions. First, we introduce ParamExplorer, an interactive and modular framework inspired by reinforcement learning that helps the exploration of parameter spaces in generative art algorithms, guided by human-in-the-loop or even automated feedback. The framework also integrates seamlessly with existing p5.js projects. Second, within this framework we implement and evaluate several exploration strategies, referred to as agents.

Problem

• Explores high-dimensional parameter spaces in generative art
• Reduces manual trial-and-error for discovering aesthetic outputs
• Provides interactive framework with human or automated feedback

Innovation

• Interactive framework for exploring generative art parameters
• Modular system with reinforcement learning-inspired exploration strategies
• Seamless integration with existing p5.js projects