Infinite Nature: Perpetual View Generation of Natural Scenes from a Single Image

TL;DR

This paper presents a hybrid method for long-range view generation from a single image, combining geometry and image synthesis to produce plausible natural scene sequences over extended camera trajectories.

Contribution

It introduces a novel iterative framework that integrates geometry and image synthesis for perpetual view generation from a single image, enabling long-range scene synthesis.

Findings

Outperforms existing methods in generating longer scene sequences

Can be trained on monocular video data

Effective on aerial coastal scene footage

Abstract

We introduce the problem of perpetual view generation - long-range generation of novel views corresponding to an arbitrarily long camera trajectory given a single image. This is a challenging problem that goes far beyond the capabilities of current view synthesis methods, which quickly degenerate when presented with large camera motions. Methods for video generation also have limited ability to produce long sequences and are often agnostic to scene geometry. We take a hybrid approach that integrates both geometry and image synthesis in an iterative `\emph{render}, \emph{refine} and \emph{repeat}' framework, allowing for long-range generation that cover large distances after hundreds of frames. Our approach can be trained from a set of monocular video sequences. We propose a dataset of aerial footage of coastal scenes, and compare our method with recent view synthesis and conditional video generation baselines, showing that it can generate plausible scenes for much longer time horizons over large camera trajectories compared to existing methods. Project page at https://infinite-nature.github.io/.