Human motion generation stands as a significant pursuit in generative computer vision, while achieving long-sequence and efficient motion generation remains challenging. Recent advancements in state space models (SSMs), notably Mamba, have showcased considerable promise in long sequence modeling with an efficient hardware-aware design, which appears to be a promising direction to build motion generation model upon it. Nevertheless, adapting SSMs to motion generation faces hurdles since the lack of a specialized design architecture to model motion sequence. To address these challenges, we propose Motion Mamba, a simple and efficient approach that presents the pioneering motion generation model utilized SSMs. Specifically, we design a Hierarchical Temporal Mamba (HTM) block to process temporal data by ensemble varying numbers of isolated SSM modules across a symmetric U-Net architecture aimed at preserving motion consistency between frames. We also design a Bidirectional Spatial Mamba (BSM) block to bidirectionally process latent poses, to enhance accurate motion generation within a temporal frame. Our proposed method achieves up to 50% FID improvement and up to 4 times faster on the HumanML3D and KIT-ML datasets compared to the previous best diffusion-based method, which demonstrates strong capabilities of high-quality long sequence motion modeling and real-time human motion generation.

News

(3/15/2024) 🎉 Our paper has been promoted by MarkTechPost!

(3/13/2024) 🎉 Our paper has been featured in Daily Papers!

(3/13/2024) 🎉 Our paper has been promoted by CVer!

Citation

@article{zhang2024motion,
  title={Motion Mamba: Efficient and Long Sequence Motion Generation with Hierarchical and Bidirectional Selective SSM},
  author={Zhang, Zeyu and Liu, Akide and Reid, Ian and Hartley, Richard and Zhuang, Bohan and Tang, Hao},
  journal={arXiv preprint arXiv:2403.07487},
  year={2024}
}

Acknowledgements

• Mamba: Linear-Time Sequence Modeling with Selective State Spaces

• Motion Latent Diffusion: Executing your Commands via Motion Diffusion in Latent Space