FlowMotion: Target-Predictive Conditional Flow Matching for Jitter-Reduced Text-Driven Human Motion Generation
This work addresses the problem of jitter in text-driven human motion generation for applications requiring smooth animations, representing an incremental improvement over existing flow-matching methods.
The paper tackled the challenge of generating high-fidelity and temporally smooth 3D human motion in resource-constrained environments by introducing FlowMotion, a method based on Conditional Flow Matching that improves target motion prediction, resulting in state-of-the-art jitter reduction on the KIT dataset and competitive FID scores.
Achieving high-fidelity and temporally smooth 3D human motion generation remains a challenge, particularly within resource-constrained environments. We introduce FlowMotion, a novel method leveraging Conditional Flow Matching (CFM). FlowMotion incorporates a training objective within CFM that focuses on more accurately predicting target motion in 3D human motion generation, resulting in enhanced generation fidelity and temporal smoothness while maintaining the fast synthesis times characteristic of flow-matching-based methods. FlowMotion achieves state-of-the-art jitter performance, achieving the best jitter in the KIT dataset and the second-best jitter in the HumanML3D dataset, and a competitive FID value in both datasets. This combination provides robust and natural motion sequences, offering a promising equilibrium between generation quality and temporal naturalness.