Hierarchical Attention and Graph Neural Networks: Toward Drift-Free Pose Estimation

TL;DR

This paper introduces a learned model using hierarchical attention and graph neural networks to improve 3D pose estimation accuracy, reducing drift and enhancing rotational component accuracy compared to traditional methods.

Contribution

It replaces traditional geometric registration and pose graph optimization with a novel learned framework that condenses data flow for precise rigid pose estimation.

Findings

Significant improvement in pose accuracy on KITTI dataset

Enhanced rotational component estimation over traditional methods

Effective reduction of drift in pose estimation

Abstract

The most commonly used method for addressing 3D geometric registration is the iterative closet-point algorithm, this approach is incremental and prone to drift over multiple consecutive frames. The Common strategy to address the drift is the pose graph optimization subsequent to frame-to-frame registration, incorporating a loop closure process that identifies previously visited places. In this paper, we explore a framework that replaces traditional geometric registration and pose graph optimization with a learned model utilizing hierarchical attention mechanisms and graph neural networks. We propose a strategy to condense the data flow, preserving essential information required for the precise estimation of rigid poses. Our results, derived from tests on the KITTI Odometry dataset, demonstrate a significant improvement in pose estimation accuracy. This improvement is especially notable in determining rotational components when compared with results obtained through conventional multi-way registration via pose graph optimization. The code will be made available upon completion of the review process.