{"id":"arxiv/2607.07306","url":"https://arxiv.org/abs/2607.07306","tags":["cs.RO","eess.SY","math.OC"],"score":0,"title":"Wrench-Based Bayesian Pose Estimation via Matrix--Fisher Gaussian Inference","domain":"cs.RO","source":"arxiv","authors":["Jianyu Chen","Lin Yang","Yibang Li","Domenico Campolo","Cyrus Mostajeran"],"reviews":0,"subject":"cs.RO","summary":"In this paper, a residual-safeguarded local Matrix Fisher--Gaussian (MFG) inference method is developed for wrench-based pose estimation on $\\mathrm{SO}(3)\\times\\mathbb{R}^3$. The force/torque measurements are modeled by a quasi-static contact system in which the predicted wrench depends on the unknown object pose through an implicit equilibrium state. Since the resulting nonlinear likelihood is not globally conjugate to the coupled MFG family, a local Bayesian update is constructed by linearizing the reduced wrench residual and matching the induced Gauss--Newton posterior model to a coupled…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-07-08","reactions":{}}