Hybrid Localization using Model- and Learning-Based Methods: Fusion of Monte Carlo and E2E Localizations via Importance Sampling

Abstract

This paper presents an efficient solution to 3D-LiDAR-based Monte Carlo localization (MCL). MCL robustly works if particles are exactly sampled around the ground truth. An inertial navigation system (INS) can be used for accurate sampling, but many particles are still needed to be used for solving the 3D localization problem even if INS is available. In particular, huge number of particles are necessary if INS is not available and it makes infeasible to perform 3D MCL in terms of the computational cost. Scan matching (SM), that is optimization-based localization, efficiently works even though INS is not available because SM can ignore movement constraints of a robot and/or device in its optimization process. However, SM sometimes determines an infeasible estimate against movement. We consider that MCL and SM have complemental advantages and disadvantages and propose a fusion method of MCL and SM. Because SM is considered as optimization of a measurement model in terms of the probabilistic modeling, we perform measurement model optimization as SM. The optimization result is then used to approximate the measurement model distribution and the approximated distribution is used to sample particles. The sampled particles are fused with MCL via importance sampling. As a result, the advantages of MCL and SM can be simultaneously utilized while mitigating their disadvantages. Experiments are conducted on the KITTI dataset and other two open datasets. Results show that the presented method can be run on a single CPU thread and accurately perform localization even if INS is not available.

NDT-Transformer: Large-Scale 3D Point Cloud Localisation using the Normal Distribution Transform Representation

“…The state-of-the-art point cloud based localisation methods can be categorised into three main streams: regression-based [1], [2], [3], intermediate-representation-based [4], [5], [6] and global-feature-based [9], [10], [11], [12], [13], [14], [15].…”

“…Following the success of deep pose estimation in imagebased global localisation, some methods [1], [2], [3] propose to use deep regression networks to learn the 6 DOF global pose. These model-based approaches are extremely efficient but not generalisable.…”

“…The core challenges for current robot localisation systems are generality and scalability. For example, model-based methods [1], [2], [3] can learn the 6 DOF global pose as a regression model and infer the pose in real-time. Sadly, regression-based methods cannot be generalised to novel environments.…”

TreeLoc: An Ensemble Learning-based Approach for Range Based Indoor Localization

“…An ensemble method is a technique that combines the predictions from multiple machine learning algorithms together to make more accurate predictions than any individual model. Few works existing on applying ensemble learning to indoor localization problems [20][21][22][23]. In general boosting, it uses a weighted averaging principle to calculate the final prediction [20].…”

Asynchronous Collaborative Localization by Integrating Spatiotemporal Graph Learning with Model-Based Estimation

“…Recently, several hybrid approaches were proposed for state estimation by integrating model-based and learningbased methods, e.g., by integrating convolutional neural networks and Bayesian filters for robot pose estimation [25], [26], and hybrid sensor fusion based on Gaussian process for location query [27]. However, existing hybrid approaches cannot well address asynchronous observations caused by the delay of robot communications.…”