EMDM:用于快速高质量运动生成的高效运动扩散模型
EMDM: Efficient Motion Diffusion Model for Fast and High-Quality Motion Generation
AI总结:
本文提出高效运动扩散模型 EMDM,通过条件去噪扩散 GAN 在少步采样下捕捉复杂多模态分布并结合几何损失,实现实时且高质量的人体运动生成。
摘要:
我们提出高效运动扩散模型(EMDM),用于快速且高质量的人体运动生成。当前最先进的生成式扩散模型已经取得了令人印象深刻的结果,但难以在不牺牲质量的情况下实现快速生成。一方面,先前的工作(如运动潜在扩散)为了提高效率而在潜在空间中进行扩散,但学习这样的潜在空间可能是一项不小的努力。另一方面,通过简单地增大采样步长(例如 DDIM)来加速生成,往往会导致质量下降,因为它无法近似复杂的去噪分布。为了解决这些问题,我们提出 EMDM,它在扩散模型的多个采样步骤中捕捉复杂分布,从而允许大幅减少采样步骤并显著加速生成。这是通过一个条件去噪扩散 GAN 来实现的,该 GAN 在任意(且可能更大的)步长下、以控制信号为条件捕捉多模态数据分布,从而能够以高保真度和多样性进行少步运动采样。为了尽量减少不希望出现的运动伪影,在网络学习过程中施加了几何损失。因此,EMDM 实现了实时运动生成,与现有方法相比显著提高了运动扩散模型的效率,同时实现了高质量的运动生成。我们的代码将在发表后公开。
英文摘要:
We introduce Efficient Motion Diffusion Model (EMDM) for fast and high-quality human motion generation. Current state-of-the-art generative diffusion models have produced impressive results but struggle to achieve fast generation without sacrificing quality. On the one hand, previous works, like motion latent diffusion, conduct diffusion within a latent space for efficiency, but learning such a latent space can be a non-trivial effort. On the other hand, accelerating generation by naively increasing the sampling step size, e.g., DDIM, often leads to quality degradation as it fails to approximate the complex denoising distribution. To address these issues, we propose EMDM, which captures the complex distribution during multiple sampling steps in the diffusion model, allowing for much fewer sampling steps and significant acceleration in generation. This is achieved by a conditional denoising diffusion GAN to capture multimodal data distributions among arbitrary (and potentially larger) step sizes conditioned on control signals, enabling fewer-step motion sampling with high fidelity and diversity. To minimize undesired motion artifacts, geometric losses are imposed during network learning. As a result, EMDM achieves real-time motion generation and significantly improves the efficiency of motion diffusion models compared to existing methods while achieving high-quality motion generation. Our code will be publicly available upon publication.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。