We present a method for learning to generate unbounded flythrough videos of natural scenes starting from a single view, where this capability is learned from a collection of single photographs, without requiring camera poses or even multiple views of each scene. To achieve this, we propose a novel self-supervised view generation training paradigm, where we sample and rendering virtual camera trajectories, including cyclic ones, allowing our model to learn stable view generation from a collection of single views. At test time, despite never seeing a video during training, our approach can take a single image and generate long camera trajectories comprised of hundreds of new views with realistic and diverse content. We compare our approach with recent state-of-the-art supervised view generation methods that require posed multi-view videos and demonstrate superior performance and synthesis quality.
@article{arxiv.2207.11148,
title = {InfiniteNature-Zero: Learning Perpetual View Generation of Natural Scenes from Single Images},
author = {Zhengqi Li and Qianqian Wang and Noah Snavely and Angjoo Kanazawa},
journal= {arXiv preprint arXiv:2207.11148},
year = {2022}
}