vMF-Contact: Uncertainty-aware Evidential Learning for Probabilistic Contact-grasp in Noisy Clutter

TL;DR

This paper introduces vMF-Contact, a novel probabilistic learning architecture that captures both aleatoric and epistemic uncertainties for robust 6-DoF grasp detection in noisy, cluttered environments, with formal guarantees and improved real-world performance.

Contribution

The paper presents vMF-Contact, a new hierarchical contact grasp model using von Mises-Fisher distribution for uncertainty quantification, enhancing grasp prediction in noisy settings.

Findings

39% improvement in clearance rate over baselines

Effective modeling of directional uncertainty with vMF distribution

Enhanced generalization to unseen objects

Abstract

Grasp learning in noisy environments, such as occlusions, sensor noise, and out-of-distribution (OOD) objects, poses significant challenges. Recent learning-based approaches focus primarily on capturing aleatoric uncertainty from inherent data noise. The epistemic uncertainty, which represents the OOD recognition, is often addressed by ensembles with multiple forward paths, limiting real-time application. In this paper, we propose an uncertainty-aware approach for 6-DoF grasp detection using evidential learning to comprehensively capture both uncertainties in real-world robotic grasping. As a key contribution, we introduce vMF-Contact, a novel architecture for learning hierarchical contact grasp representations with probabilistic modeling of directional uncertainty as von Mises-Fisher (vMF) distribution. To achieve this, we analyze the theoretical formulation of the second-order objective on the posterior parametrization, providing formal guarantees for the model's ability to quantify uncertainty and improve grasp prediction performance. Moreover, we enhance feature expressiveness by applying partial point reconstructions as an auxiliary task, improving the comprehension of uncertainty quantification as well as the generalization to unseen objects. In the real-world experiments, our method demonstrates a significant improvement by 39% in the overall clearance rate compared to the baselines. The code is available under: https://github.com/YitianShi/vMF-Contact/