Artists and video game designers often construct 2D animations using libraries of sprites -- textured patches of objects and characters. We propose a deep learning approach that decomposes sprite-based video animations into a disentangled representation of recurring graphic elements in a self-supervised manner. By jointly learning a dictionary of possibly transparent patches and training a network that places them onto a canvas, we deconstruct sprite-based content into a sparse, consistent, and explicit representation that can be easily used in downstream tasks, like editing or analysis. Our framework offers a promising approach for discovering recurring visual patterns in image collections without supervision.
@article{arxiv.2104.14553,
title = {MarioNette: Self-Supervised Sprite Learning},
author = {Dmitriy Smirnov and Michael Gharbi and Matthew Fisher and Vitor Guizilini and Alexei A. Efros and Justin Solomon},
journal= {arXiv preprint arXiv:2104.14553},
year = {2021}
}