Stereo Magnification: Learning View Synthesis using Multiplane Images

TL;DR

This paper introduces a deep learning approach using multiplane images to synthesize novel views from stereo images, enabling significant view extrapolation and applications in magnifying narrow-baseline stereo cameras.

Contribution

We propose a new layered representation called multiplane images and a learning framework trained on YouTube videos for view synthesis from stereo pairs.

Findings

Our method outperforms recent view synthesis techniques.

It can extrapolate views beyond the input baseline.

Demonstrates applications in magnifying stereo images.

Abstract

The view synthesis problem--generating novel views of a scene from known imagery--has garnered recent attention due in part to compelling applications in virtual and augmented reality. In this paper, we explore an intriguing scenario for view synthesis: extrapolating views from imagery captured by narrow-baseline stereo cameras, including VR cameras and now-widespread dual-lens camera phones. We call this problem stereo magnification, and propose a learning framework that leverages a new layered representation that we call multiplane images (MPIs). Our method also uses a massive new data source for learning view extrapolation: online videos on YouTube. Using data mined from such videos, we train a deep network that predicts an MPI from an input stereo image pair. This inferred MPI can then be used to synthesize a range of novel views of the scene, including views that extrapolate significantly beyond the input baseline. We show that our method compares favorably with several recent view synthesis methods, and demonstrate applications in magnifying narrow-baseline stereo images.