4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time

TL;DR

4D-LRM is a large-scale, unified 4D reconstruction model capable of rendering arbitrary view-time combinations efficiently and accurately, demonstrating strong generalization and interpolation abilities across diverse scenarios.

Contribution

It introduces the first scalable 4D reconstruction model that learns a unified space-time representation from unconstrained views and timestamps, enabling fast and high-quality rendering.

Findings

Reconstructs 24-frame sequences in under 1.5 seconds on a single GPU.

Generalizes well to novel objects and diverse camera setups.

Effectively interpolates across time and view angles.

Abstract

Can we scale 4D pretraining to learn general space-time representations that reconstruct an object from a few views at some times to any view at any time? We provide an affirmative answer with 4D-LRM, the first large-scale 4D reconstruction model that takes input from unconstrained views and timestamps and renders arbitrary novel view-time combinations. Unlike prior 4D approaches, e.g., optimization-based, geometry-based, or generative, that struggle with efficiency, generalization, or faithfulness, 4D-LRM learns a unified space-time representation and directly predicts per-pixel 4D Gaussian primitives from posed image tokens across time, enabling fast, high-quality rendering at, in principle, infinite frame rate. Our results demonstrate that scaling spatiotemporal pretraining enables accurate and efficient 4D reconstruction. We show that 4D-LRM generalizes to novel objects, interpolates across time, and handles diverse camera setups. It reconstructs 24-frame sequences in one forward pass with less than 1.5 seconds on a single A100 GPU.