融合视觉触觉与仿真推断，实现高精度插孔操作的不确定姿态估计。
BayesContact: Uncertain Pose Estimation via Visuo-Tactile Proposals and Simulation-based Inference
- 用仿真模拟生成深度与接触信号，实时更新物体姿态信念
- 在模拟与真实机器人上使插入成功率提升30%
- 适合需要高精度定位的复杂抓取任务
高接触场景下的操作任务对姿态估计精度要求极高，远超仅靠深度感知的能力。现有方法依赖视觉与触觉信息，但需昂贵的离线训练，且无法适应新环境和几何形状。本文提出BayesContact，一种基于仿真的推断框架，用于执行插孔插入任务中的多模态姿态估计。该方法维护物体姿态的粒子信念，融合深度观测与力/力矩衍生的接触证据。通过仿真前向模型近似观测似然：对每个姿态假设，渲染器预测深度图，物理模拟器预测受控探测动作下的接触结果；两者与真实观测对比后用于更新信念。所获多模态信念支持基于信息增益的主动探测以消除歧义。在多种模拟几何与真实机器人实验中，相比纯视觉推断，其姿态可观测性与插入成功率提升30%。
原文摘要 · Abstract (English)
Contact-rich manipulation requires pose estimates that are often more accurate than what depth-only sensing provides. Existing methods, relying on vision and contact, employ costly offline training procedures that need to be retrained for new environments and geometries. We propose BayesContact, a Simulation-Based Inference framework for visuo-tactile pose estimation in peg-in-hole insertion. BayesContact maintains a particle belief over object pose and fuses depth observations with force/torque-derived contact evidence. We employ simulation based forward models to approximate these observation likelihoods. For each pose hypothesis, a renderer predicts depth measurements and a physics simulator predicts contact outcomes under guarded probing actions; both are scored against real observations to update the belief. The resulting multimodal belief also enables information-gain-based probing for active disambiguation. Across simulated geometries and real-robot experiments, BayesContact improves pose observability and insertion success over vision-only inference by 30%
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。