FlowMotion: Training-Free Flow Guidance for Video Motion Transfer

TL;DR

This paper presents FlowMotion, a novel training-free framework that enables efficient and flexible motion transfer by directly leveraging the predicted outputs of flow-based T2V models and proposes flow guidance, which extracts motion representations based on latent predictions to align motion patterns between source and generated videos.

Abstract

Video motion transfer aims to generate a target video that inherits motion patterns from a source video while rendering new scenes. Existing training-free approaches focus on constructing motion guidance based on the intermediate outputs of pre-trained T2V models, which results in heavy computational overhead and limited flexibility. In this paper, we present FlowMotion, a novel training-free framework that enables efficient and flexible motion transfer by directly leveraging the predicted outputs of flow-based T2V models. Our key insight is that early latent predictions inherently encode rich temporal information. Motivated by this, we propose flow guidance, which extracts motion representations based on latent predictions to align motion patterns between source and generated videos. We further introduce a velocity regularization strategy to stabilize optimization and ensure smooth motion evolution. By operating purely on model predictions, FlowMotion achieves superior time and resource efficiency as well as competitive performance compared with state-of-the-art methods.

Figures (10)

• Figure 1: Guidance process of exisiting methods and FlowMotion.
• Figure 2: Qualitative comparison with SOTA methods.Red boxes indicated low quality content across frames.
• Figure 3: Qualitative results of FlowMotion. The left column generated by Wan2.1-1.3B, while right column generated by Wan2.2-5B.
• Figure 4: The visualization of ablation on key designs. Red boxes indicate low quality content across frames.
• Figure 5: Variance of different source motion representation.