Motion-Zero: Zero-Shot Moving Object Control Framework for Diffusion-Based Video Generation

TL;DR

Motion-Zero is a zero-shot framework that enables precise control of moving object trajectories in diffusion-based video generation, ensuring spatial and temporal consistency without additional training.

Contribution

It introduces a novel zero-shot control method with spatial and temporal constraints for diffusion models, enhancing motion accuracy and video quality.

Findings

Effective trajectory control of objects in generated videos

Maintains high visual quality and spatial-temporal consistency

Applicable to various diffusion models without retraining

Abstract

Recent large-scale pre-trained diffusion models have demonstrated a powerful generative ability to produce high-quality videos from detailed text descriptions. However, exerting control over the motion of objects in videos generated by any video diffusion model is a challenging problem. In this paper, we propose a novel zero-shot moving object trajectory control framework, Motion-Zero, to enable a bounding-box-trajectories-controlled text-to-video diffusion model. To this end, an initial noise prior module is designed to provide a position-based prior to improve the stability of the appearance of the moving object and the accuracy of position. In addition, based on the attention map of the U-net, spatial constraints are directly applied to the denoising process of diffusion models, which further ensures the positional and spatial consistency of moving objects during the inference. Furthermore, temporal consistency is guaranteed with a proposed shift temporal attention mechanism. Our method can be flexibly applied to various state-of-the-art video diffusion models without any training process. Extensive experiments demonstrate our proposed method can control the motion trajectories of objects and generate high-quality videos. Our project page is https://vpx-ecnu.github.io/MotionZero-website/