LatentSLAM: unsupervised multi-sensor representation learning for localization and mapping

Biologically inspired algorithms for simultaneous localization and mapping (SLAM) such as RatSLAM have been shown to yield effective and robust robot navigation in both indoor and outdoor environments. One drawback however is the sensitivity to perceptual aliasing due to the template matching of low-dimensional sensory templates. In this paper, we propose an unsupervised representation learning method that yields low-dimensional latent state descriptors that can be used for RatSLAM. Our method is sensor agnostic and can be applied to any sensor modality, as we illustrate for camera images, radar range-doppler maps and lidar scans. We also show how combining multiple sensors can increase the robustness, by reducing the number of false matches. We evaluate on a dataset captured with a mobile robot navigating in a warehouse-like environment, moving through different aisles with similar appearance, making it hard for the SLAM algorithms to disambiguate locations.

[1]  Stefan Wermter,et al.  RatSLAM on Humanoids - A Bio-Inspired SLAM Model Adapted to a Humanoid Robot , 2014, ICANN.

[2]  Danilo Jimenez Rezende,et al.  Generalized ELBO with Constrained Optimization , GECO , 2018 .

[3]  Wei Yang,et al.  Are We Ready for Service Robots? The OpenLORIS-Scene Datasets for Lifelong SLAM , 2020, 2020 IEEE International Conference on Robotics and Automation (ICRA).

[4]  Max Welling,et al.  Auto-Encoding Variational Bayes , 2013, ICLR.

[5]  Karl J. Friston.  Life as we know it , 2013, Journal of The Royal Society Interface.

[6]  Rui Yan,et al.  An Improved Loop Closure Detection for RatSLAM , 2019, 2019 5th International Conference on Control, Automation and Robotics (ICCAR).

[7]  Gordon Wyeth,et al.  RatSLAM: Using Models of Rodent Hippocampus for Robot Navigation and Beyond , 2013, ISRR.

[8]  Alessandro Treves,et al.  Attractor neural networks storing multiple space representations: A model for hippocampal place fields , 1998, cond-mat/9807101.

[9]  Yoshua Bengio,et al.  Série Scientifique Scientific Series Incorporating Second-order Functional Knowledge for Better Option Pricing Incorporating Second-order Functional Knowledge for Better Option Pricing , 2022 .

[10]  J. M. M. Montiel,et al.  ORB-SLAM: A Versatile and Accurate Monocular SLAM System , 2015, IEEE Transactions on Robotics.

[11]  Gordon Wyeth,et al.  FAB-MAP + RatSLAM: Appearance-based SLAM for multiple times of day , 2010, 2010 IEEE International Conference on Robotics and Automation.

[12]  Tim Verbelen,et al.  Learning Perception and Planning With Deep Active Inference , 2020, ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[13]  Andrew L. Maas.  Rectifier Nonlinearities Improve Neural Network Acoustic Models , 2013 .

[14]  Kshitij Tiwari,et al.  ViTa-SLAM: A Bio-inspired Visuo-Tactile SLAM for Navigation while Interacting with Aliased Environments , 2019, 2019 IEEE International Conference on Cyborg and Bionic Systems (CBS).

[15]  Gordon Wyeth,et al.  Persistent Navigation and Mapping using a Biologically Inspired SLAM System , 2010, Int. J. Robotics Res..

[16]  Michael Milford,et al.  Towards Brain-based Sensor Fusion for Navigating Robots , 2012, ICRA 2012.

[17]  Tim Verbelen,et al.  Learning Generative State Space Models for Active Inference , 2020, Frontiers in Computational Neuroscience.

[18]  Gordon Wyeth,et al.  RatSLAM: a hippocampal model for simultaneous localization and mapping , 2004, IEEE International Conference on Robotics and Automation, 2004. Proceedings. ICRA '04. 2004.

[19]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[20]  Sam Leroux,et al.  Anomaly Detection for Autonomous Guided Vehicles using Bayesian Surprise , 2020, 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[21]  John J. Leonard,et al.  Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age , 2016, IEEE Transactions on Robotics.

[22]  François Michaud,et al.  RTAB‐Map as an open‐source lidar and visual simultaneous localization and mapping library for large‐scale and long‐term online operation , 2018, J. Field Robotics.

[23]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[24]  Jan Steckel,et al.  BatSLAM: Simultaneous Localization and Mapping Using Biomimetic Sonar , 2013, PloS one.

[25]  Lining Sun,et al.  Robustness Improvement of Visual Templates Matching Based on Frequency-Tuned Model in RatSLAM , 2020, Frontiers in Neurorobotics.

[26]  Michael Milford,et al.  Deep learning features at scale for visual place recognition , 2017, 2017 IEEE International Conference on Robotics and Automation (ICRA).

[27]  Mathew H. Evans,et al.  Tactile SLAM with a biomimetic whiskered robot , 2012, 2012 IEEE International Conference on Robotics and Automation.

[28]  Janet Wiles,et al.  OpenRatSLAM: an open source brain-based SLAM system , 2013, Autonomous Robots.

[29]  Quoc V. Le,et al.  Do Better ImageNet Models Transfer Better? , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[30]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).