LaVR: Scene Latent Conditioned Generative Video Trajectory Re-Rendering using Large 4D Reconstruction Models

TL;DR

This paper introduces LaVR, a novel method for scene-aware video re-rendering from monocular videos using large 4D reconstruction models to improve view synthesis accuracy.

Contribution

LaVR leverages implicit scene geometry in latent space of large 4D models to enhance video re-rendering, overcoming limitations of existing geometry-dependent methods.

Findings

Achieves state-of-the-art results on video re-rendering tasks.

Utilizes latent space for flexible scene representation without explicit reconstruction.

Regularizes errors effectively with pretrained diffusion prior.

Abstract

Given a monocular video, the goal of video re-rendering is to generate views of the scene from a novel camera trajectory. Existing methods face two distinct challenges. Geometrically unconditioned models lack spatial awareness, leading to drift and deformation under viewpoint changes. On the other hand, geometrically-conditioned models depend on estimated depth and explicit reconstruction, making them susceptible to depth inaccuracies and calibration errors. We propose to address these challenges by using the implicit geometric knowledge embedded in the latent space of a large 4D reconstruction model to condition the video generation process. These latents capture scene structure in a continuous space without explicit reconstruction. Therefore, they provide a flexible representation that allows the pretrained diffusion prior to regularize errors more effectively. By jointly conditioning on these latents and source camera poses, we demonstrate that our model achieves state-of-the-art results on the video re-rendering task. Project webpage is https://lavr-4d-scene-rerender.github.io/.