Point Cloud based Hierarchical Deep Odometry Estimation

TL;DR

This paper introduces a hierarchical deep learning model that estimates vehicle odometry directly from raw point cloud data, with a novel local bundle adjustment approach using LSTM layers, evaluated against current state-of-the-art methods.

Contribution

It presents a new hierarchical deep neural network architecture for odometry estimation from point clouds, including a local bundle adjustment variation with LSTM layers, advancing the state-of-the-art.

Findings

The hierarchical model effectively estimates odometry from raw point clouds.

The local bundle adjustment with LSTM improves estimation accuracy.

The proposed methods outperform existing approaches in evaluations.

Abstract

Processing point clouds using deep neural networks is still a challenging task. Most existing models focus on object detection and registration with deep neural networks using point clouds. In this paper, we propose a deep model that learns to estimate odometry in driving scenarios using point cloud data. The proposed model consumes raw point clouds in order to extract frame-to-frame odometry estimation through a hierarchical model architecture. Also, a local bundle adjustment variation of this model using LSTM layers is implemented. These two approaches are comprehensively evaluated and are compared against the state-of-the-art.