Abstract
This article suggests a reasoning-guided vision-language-motion diffusion framework (RG-VLMD) for generating instruction-aware co-speech gestures for humanoid robots in educational scenarios. The system integrates multi-modal affective estimation, pedagogical reasoning, and teaching-act-conditioned motion synthesis to enable adaptive and semantically consistent robot behavior. A gated mixture-of-experts model predicts Valence/Arousal from input text, visual, and acoustic features, which then mapped to discrete teaching-act categories through an affect-driven this http URL signals condition a diffusion-based motion generator using clip-level intent and frame-level instructional schedules via additive latent restriction with auxiliary action-group supervision. Compared to a baseline diffusion model, our proposed method produces more structured and distinctive motion patterns, as verified by motion statics and pairwise distance analysis. Generated motion sequences remain physically plausible and can be retargeted to a NAO robot for real-time execution. The results reveal that reasoning-guided instructional conditioning improves gesture controllability and pedagogical expressiveness in educational human-robot interaction.
Abstract (translated)
本文提出了一种推理引导的视觉-语言-运动扩散框架（RG-VLMD），用于为教育场景中的人形机器人生成指令感知的共讲手势。该系统整合了多模态情感估计、教学推理和教学行为条件运动合成，以实现自适应且语义一致的机器人行为。一个门控专家混合模型从输入的文本、视觉和声学特征中预测效价/唤醒度，随后通过情感驱动信号将其映射到离散的教学行为类别。这些信号通过添加式潜在约束和辅助动作组监督，以片段级意图和帧级教学调度来条件化一个基于扩散的运动生成器。与基线扩散模型相比，本文方法产生了更具结构性和差异性的运动模式，这通过运动统计量和成对距离分析得到了验证。生成的运动序列保持物理合理性，并可重新定位到NAO机器人上实现实时执行。结果表明，推理引导的教学条件化提高了教育人机交互中的手势可控性和教学表现力。
URL
https://arxiv.org/abs/2603.18771
