Forming a sparse representation for visual place recognition using a neurorobotic approach

This paper introduces a novel unsupervised neural network model for visual information encoding which aims to address the problem of large-scale visual localization. Inspired by the structure of the visual cortex, the model (namely HSD) alternates layers of topologic sparse coding and pooling to build a more compact code of visual information. Intended for visual place recognition (VPR) systems that use local descriptors, the impact of its integration in a bio-inpired model for self-localization (LPMP) is evaluated. Our experimental results on the KITTI dataset show that HSD improves the runtime speed of LPMP by a factor of at least 2 and its localization accuracy by 10%. A comparison with CoHog, a state-of-the-art VPR approach, showed that our method achieves slightly better results.

[1]  Eduard Kuriscak,et al.  Biological context of Hebb learning in artificial neural networks, a review , 2015, Neurocomputing.

[2]  Homayoun Najjaran,et al.  Autonomous vehicle perception: The technology of today and tomorrow , 2018 .

[3]  Thomas Serre,et al.  Robust Object Recognition with Cortex-Like Mechanisms , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Peter I. Corke,et al.  Visual Place Recognition: A Survey , 2016, IEEE Transactions on Robotics.

[5]  Philippe Gaussier,et al.  Neurobiologically Inspired Mobile Robot Navigation and Planning , 2007, Frontiers in neurorobotics.

[6]  Alexander Carballo,et al.  A Survey of Autonomous Driving: Common Practices and Emerging Technologies , 2019, IEEE Access.

[7]  Dario L. Ringach,et al.  Link between orientation and retinotopic maps in primary visual cortex , 2012, Proceedings of the National Academy of Sciences.

[8]  Stéphane Mallat,et al.  Matching pursuits with time-frequency dictionaries , 1993, IEEE Trans. Signal Process..

[9]  Olivier Romain,et al.  Position Paper: Prototyping Autonomous Vehicles Applications with Heterogeneous Multi-FpgaSystems , 2019, 2019 UK/ China Emerging Technologies (UCET).

[10]  Andreas Geiger,et al.  Vision meets robotics: The KITTI dataset , 2013, Int. J. Robotics Res..

[11]  M. V. Velzen,et al.  Self-organizing maps , 2007 .

[12]  H. Jörntell,et al.  Questioning the role of sparse coding in the brain , 2015, Trends in Neurosciences.

[13]  Michael Milford,et al.  CoHOG: A Light-Weight, Compute-Efficient, and Training-Free Visual Place Recognition Technique for Changing Environments , 2020, IEEE Robotics and Automation Letters.

[14]  Sebastien Glaser,et al.  Simultaneous Localization and Mapping: A Survey of Current Trends in Autonomous Driving , 2017, IEEE Transactions on Intelligent Vehicles.

[15]  Philippe Gaussier,et al.  Orientation system in Robots: Merging Allothetic and Idiothetic Estimations , 2007 .

[16]  Paul Newman,et al.  1 year, 1000 km: The Oxford RobotCar dataset , 2017, Int. J. Robotics Res..

[17]  Michael Milford,et al.  VPR-Bench: An Open-Source Visual Place Recognition Evaluation Framework with Quantifiable Viewpoint and Appearance Change , 2020, International Journal of Computer Vision.

[18]  Laurent U. Perrinet,et al.  Role of Homeostasis in Learning Sparse Representations , 2007, Neural Computation.

[19]  Sarosh H. Patel,et al.  Review of Neurobiologically Based Mobile Robot Navigation System Research Performed Since 2000 , 2016, J. Robotics.

[20]  Nikolaos Doulamis,et al.  Deep Learning for Computer Vision: A Brief Review , 2018, Comput. Intell. Neurosci..

[21]  Olivier Romain,et al.  From Neurorobotic Localization to Autonomous Vehicles , 2019, Unmanned Syst..

[22]  R. Reid,et al.  Rules of Connectivity between Geniculate Cells and Simple Cells in Cat Primary Visual Cortex , 2001, The Journal of Neuroscience.

[23]  Bruno A Olshausen,et al.  Sparse coding of sensory inputs , 2004, Current Opinion in Neurobiology.

[24]  Xiang Zhang,et al.  OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks , 2013, ICLR.