无需训练，用提示词让多模态大模型精准理解微动作背后的身心状态。
Recognition-Conditioned Reasoning: A Training-Free Multimodal-LLM Pipeline for Fine-Grained Micro-Action Understanding
- 按任务类型动态分配最优模型：识别用判别式，描述推理用生成式。
- 在开放任务上得分2.68（满分5），显著优于第二名的1.44。
- 适合需要零训练、高解释性的微动作分析场景。
微动作是人类无意识做出的细微、短暂且幅度小的身体动作，如手指轻颤或头部微倾，能可靠反映情绪与心理状态。理解这些动作不仅需打标签，还需描述具体动作部位并合理推断其背后原因。我们提出一种完全无需训练、仅依赖提示词的系统，在不允许微调和真实标签监督的MAC 2026微动作挑战赛细粒度理解赛道中夺冠。系统基于冻结的多模态大语言模型（MLLM），将八个子任务动态分配给最适配的模型：判别式MLLM用于封闭式识别，生成式MLLM用于开放式描述与推理。该架构在开放任务上表现显著更优，平均得分达2.68（五分制），远超第二名的1.44。
原文摘要 · Abstract (English)
Micro-actions are subtle, short, low-amplitude body movements, such as a fidgeting hand or a slight head tilt, that humans perform with little conscious intent yet that reliably leak emotional and psychological state. Understanding them goes beyond assigning a label: a model must also describe which body parts move and reason, faithfully, about why a clip warrants a particular fine-grained category. We present the training-free, prompt-only system that won first place in the fine-grained understanding track (MA-Bench) of the MAC~2026 Micro-Action Challenge, where both fine-tuning and ground-truth supervision are disallowed. Built entirely upon frozen multimodal large language models (MLLMs), the system dynamically routes each of the eight sub-tasks to the MLLM empirically best suited for that task: a discriminative MLLM for closed-ended recognition tasks and a generative MLLM for open-ended description and reasoning tasks. This architecture achieves a statistically significant performance advantage on open-ended tasks, attaining an average score of 2.68 (on a five-point scale) compared to 1.44 for the second-best approach.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。