Learning Fractals by Gradient Descent
Lists
Only you see this
Abstract
Fractals are geometric shapes that can display complex and self-similar patterns found in nature (e.g., clouds and plants). Recent works in visual recognition have leveraged this property to create random fractal images for model pre-training. In this paper, we study the inverse problem --- given a target image (not necessarily a fractal), we aim to generate a fractal image that looks like it. We propose a novel approach that learns the parameters underlying a fractal image via gradient descent. We show that our approach can find fractal parameters of high visual quality and be compatible with different loss functions, opening up several potentials, e.g., learning fractals for downstream tasks, scientific understanding, etc.
BibTeX
@inproceedings{tu2023learning,
author = {Cheng-Hao Tu and Hong-You Chen and David Carlyn and Wei-Lun Chao},
title = {{Learning Fractals by Gradient Descent}},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2023},
doi = {10.1609/aaai.v37i2.25342},
}| Learning Image Fractals Using Chaotic Differentiable Point Splatting | arXiv · 2025 |
| A Fractal Geometry Based Method for Dataset Augmentation by Automatically Generation of Multi-View Images | Journal of the Japan Society for Precision Engineering · 2021 |
| Fractal Geometry-Based Automatic Generation of Large-Scale Image Database for Pre-Training in 3D Object Recognition | Journal of the Japan Society for Precision Engineering · 2021 |
| Pre-training Without Natural Images | arXiv · 2021 |
| Formula-Supervised Visual-Geometric Pre-training | arXiv · 2024 |
| Can Vision Transformers Learn without Natural Images? | arXiv · 2021 |