SpaceTimePilot: Generative Rendering of Dynamic Scenes Across Space and Time

TL;DR

SpaceTimePilot introduces a novel generative model that independently controls space and time in dynamic scene rendering, enabling continuous exploration and manipulation of videos through a diffusion process with specialized training and datasets.

Contribution

The paper presents a new diffusion-based model with explicit space-time control, a temporal-warping training scheme, and a synthetic dataset, advancing controllable video generation from monocular inputs.

Findings

Effective space-time disentanglement demonstrated on real and synthetic data

Improved control over camera viewpoint and motion sequences

Superior results compared to prior generative video models

Abstract

We present SpaceTimePilot, a video diffusion model that disentangles space and time for controllable generative rendering. Given a monocular video, SpaceTimePilot can independently alter the camera viewpoint and the motion sequence within the generative process, re-rendering the scene for continuous and arbitrary exploration across space and time. To achieve this, we introduce an effective animation time-embedding mechanism in the diffusion process, allowing explicit control of the output video's motion sequence with respect to that of the source video. As no datasets provide paired videos of the same dynamic scene with continuous temporal variations, we propose a simple yet effective temporal-warping training scheme that repurposes existing multi-view datasets to mimic temporal differences. This strategy effectively supervises the model to learn temporal control and achieve robust space-time disentanglement. To further enhance the precision of dual control, we introduce two additional components: an improved camera-conditioning mechanism that allows altering the camera from the first frame, and CamxTime, the first synthetic space-and-time full-coverage rendering dataset that provides fully free space-time video trajectories within a scene. Joint training on the temporal-warping scheme and the CamxTime dataset yields more precise temporal control. We evaluate SpaceTimePilot on both real-world and synthetic data, demonstrating clear space-time disentanglement and strong results compared to prior work. Project page: https://zheninghuang.github.io/Space-Time-Pilot/ Code: https://github.com/ZheningHuang/spacetimepilot