Samila: A Generative Art Generator

2025-04-05

AI Summary

This study addresses the challenge of balancing stochasticity and controllability in generative art. Methodologically, it proposes a reproducible visual creation framework grounded in mathematical functions and pseudorandom mechanisms, implemented as an open-source Python library wherein all parameters are uniquely determined by two integer seeds. It introduces the concept of “visual families”: fixing one seed while varying parameterized mappings (e.g., trigonometric, fractal, or polar coordinate transformations) yields stylistically coherent yet morphologically diverse artwork sequences. The system supports configurable projection modes and 2D point-set rendering. Key contributions include: (1) the first fully deterministic, dual-seed–driven generation pipeline; (2) an interpretable mapping between seeds, functional forms, and visual semantics; and (3) strong reproducibility, pedagogical accessibility, and utility for computational aesthetics research.

Technology Category

• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Humans and AI: Game Design — Procedural Content Generation & Storytelling

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Generative art merges creativity with computation, using algorithms to produce aesthetic works. This paper introduces Samila, a Python-based generative art library that employs mathematical functions and randomness to create visually compelling compositions. The system allows users to control the generation process through random seeds, function selections, and projection modes, enabling the exploration of randomness and artistic expression. By adjusting these parameters, artists can create diverse compositions that reflect intentionality and unpredictability. We demonstrate that Samila's outputs are uniquely determined by two random generation seeds, making regeneration nearly impossible without both. Additionally, altering the point generation functions while preserving the seed produces artworks with distinct graphical characteristics, forming a visual family. Samila serves as both a creative tool for artists and an educational resource for teaching mathematical and programming concepts. It also provides a platform for research in generative design and computational aesthetics. Future developments could include AI-driven generation and aesthetic evaluation metrics to enhance creative control and accessibility.

Problem

• Generates art using algorithms and randomness
• Controls artistic output via seeds and functions
• Explores computational aesthetics and generative design

Innovation

• Python library using mathematical functions and randomness
• Controlled by random seeds and function selections
• Enables unique, seed-determined generative art