Huggbottle/noise-images-256

Noise Pattern Dataset

A dataset of synthetic noise textures generated using various procedural algorithms, designed for machine learning, procedural generation, or other purposes.

Overview

This dataset includes 8 types of noise patterns used in procedural texture generation and simulation tasks. Each sample is a grayscale 256×256 PNG image with metadata.

Noise Types

1. Perlin Noise
2. Simplex Noise
3. Worley Noise (Voronoi)
4. Fractal Noise
5. Ridge Noise
6. Cellular Noise
7. Turbulence Noise
8. Pink Noise

Each type includes multiple parameter variations and random seeds.

Dataset Structure

noise_dataset/
├── images/           # Generated noise pattern images (256x256 PNG)
├── metadata.json     # Detailed metadata for each image
└── README.md        # This file