PS6D: Point Cloud Based Symmetry-Aware 6D Object Pose Estimation in Robot Bin-Picking

TL;DR

PS6D is a novel point cloud based framework for symmetry-aware 6D object pose estimation, specifically designed for challenging industrial scenarios involving slender and multi-symmetric objects, demonstrating significant accuracy improvements and practical deployment.

Contribution

The paper introduces PS6D, a new pose estimation framework that incorporates symmetry-aware loss functions and multi-scale feature extraction for improved accuracy in industrial bin-picking.

Findings

Achieves 11.5% higher F1 score compared to state-of-the-art.

Attains 14.8% higher recall in evaluations.

Reaches 91.7% success rate in industrial bin-picking experiments.

Abstract

6D object pose estimation holds essential roles in various fields, particularly in the grasping of industrial workpieces. Given challenges like rust, high reflectivity, and absent textures, this paper introduces a point cloud based pose estimation framework (PS6D). PS6D centers on slender and multi-symmetric objects. It extracts multi-scale features through an attention-guided feature extraction module, designs a symmetry-aware rotation loss and a center distance sensitive translation loss to regress the pose of each point to the centroid of the instance, and then uses a two-stage clustering method to complete instance segmentation and pose estimation. Objects from the Sil\'eane and IPA datasets and typical workpieces from industrial practice are used to generate data and evaluate the algorithm. In comparison to the state-of-the-art approach, PS6D demonstrates an 11.5\% improvement in F$_{1_{inst}}$ and a 14.8\% improvement in Recall. The main part of PS6D has been deployed to the software of Mech-Mind, and achieves a 91.7\% success rate in bin-picking experiments, marking its application in industrial pose estimation tasks.