用单步推理实现高帧率多模态轨迹生成，解决机器人操控延迟与分布坍塌问题。
From Flow to One Step: Real-Time Multi-Modal Trajectory Policies via Implicit Maximum Likelihood Estimation-based Distribution Distillation
- 通过隐式最大似然估计蒸馏流模型，实现单次前向传播生成
- 采用双向切比雪夫距离保持多模态分布，避免轨迹平均化
- 融合多视角视觉与本体感知，支持实时重规划与动态扰动鲁棒性
基于扩散和流匹配的生成策略在机器人操作中表现优异，能建模多模态人类示范。但其依赖迭代常微分方程求解导致显著延迟，限制高频闭环控制。现有单步加速方法虽缓解开销，却常出现分布坍塌，生成平均化轨迹，无法执行连贯操作策略。本文提出一种框架，通过隐式最大似然估计（IMLE）将条件流匹配（CFM）教师模型蒸馏为快速单步学生模型。双向切比雪夫距离提供集合级目标，兼顾模式覆盖与保真度，使学生模型在单次前向传播中保留教师的多模态动作分布。统一感知编码器融合多视角RGB、深度图、点云及本体感知，生成几何感知表征。所提方法支持高频率控制，实现动态扰动下的实时滚动规划与更强鲁棒性。
原文摘要 · Abstract (English)
Generative policies based on diffusion and flow matching achieve strong performance in robotic manipulation by modeling multi-modal human demonstrations. However, their reliance on iterative Ordinary Differential Equation (ODE) integration introduces substantial latency, limiting high-frequency closed-loop control. Recent single-step acceleration methods alleviate this overhead but often exhibit distributional collapse, producing averaged trajectories that fail to execute coherent manipulation strategies. We propose a framework that distills a Conditional Flow Matching (CFM) expert into a fast single-step student via Implicit Maximum Likelihood Estimation (IMLE). A bi-directional Chamfer distance provides a set-level objective that promotes both mode coverage and fidelity, enabling preservation of the teacher multi-modal action distribution in a single forward pass. A unified perception encoder further integrates multi-view RGB, depth, point clouds, and proprioception into a geometry-aware representation. The resulting high-frequency control supports real-time receding-horizon re-planning and improved robustness under dynamic disturbances.
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。