AMD:Anatomical Motion Diffusion with Interpretable Motion Decomposition and Fusion
This work addresses the problem of synthesizing complex human motions from text for applications in animation and virtual reality, representing an incremental improvement by enhancing existing diffusion-based methods with anatomical parsing and adaptive fusion.
The paper tackles the challenge of generating realistic human motion sequences from complex or long text descriptions by proposing the Adaptable Motion Diffusion (AMD) model, which leverages a Large Language Model to parse text into interpretable anatomical scripts and uses a two-branch fusion scheme to balance text and anatomical guidance, resulting in significant outperformance over state-of-the-art models on datasets like CLCD1 and CLCD2.
Generating realistic human motion sequences from text descriptions is a challenging task that requires capturing the rich expressiveness of both natural language and human motion.Recent advances in diffusion models have enabled significant progress in human motion synthesis.However, existing methods struggle to handle text inputs that describe complex or long motions.In this paper, we propose the Adaptable Motion Diffusion (AMD) model, which leverages a Large Language Model (LLM) to parse the input text into a sequence of concise and interpretable anatomical scripts that correspond to the target motion.This process exploits the LLM's ability to provide anatomical guidance for complex motion synthesis.We then devise a two-branch fusion scheme that balances the influence of the input text and the anatomical scripts on the inverse diffusion process, which adaptively ensures the semantic fidelity and diversity of the synthesized motion.Our method can effectively handle texts with complex or long motion descriptions, where existing methods often fail. Experiments on datasets with relatively more complex motions, such as CLCD1 and CLCD2, demonstrate that our AMD significantly outperforms existing state-of-the-art models.