Active exploration using Gaussian Random Fields and Gaussian Process Implicit Surfaces

In this work we study the problem of exploring surfaces and building compact 3D representations of the environment surrounding a robot through active perception. We propose an online probabilistic framework that merges visual and tactile measurements using Gaussian Random Field and Gaussian Process Implicit Surfaces. The system investigates incomplete point clouds in order to find a small set of regions of interest which are then physically explored with a robotic arm equipped with tactile sensors. We show experimental results obtained using a PrimeSense camera, a Kinova Jaco2 robotic arm and Optoforce sensors on different scenarios. We then demostrate how to use the online framework for object detection and terrain classification.

[1]  Danica Kragic,et al.  Mind the gap - robotic grasping under incomplete observation , 2011, 2011 IEEE International Conference on Robotics and Automation.

[2]  M. Ernst,et al.  Humans integrate visual and haptic information in a statistically optimal fashion , 2002, Nature.

[3]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[4]  Anders Heyden,et al.  Reconstructing open surfaces from unorganized data points , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..

[5]  A. Ng,et al.  Touch Based Perception for Object Manipulation , 2007 .

[6]  Hugh F. Durrant-Whyte,et al.  Contextual occupancy maps using Gaussian processes , 2009, 2009 IEEE International Conference on Robotics and Automation.

[7]  Roland Siegwart,et al.  Comparison of Boosting Based Terrain Classification Using Proprioceptive and Exteroceptive Data , 2008, ISER.

[8]  Yang Gao,et al.  Deep learning for tactile understanding from visual and haptic data , 2015, 2016 IEEE International Conference on Robotics and Automation (ICRA).

[9]  Jeffrey C. Trinkle,et al.  The application of particle filtering to grasping acquisition with visual occlusion and tactile sensing , 2012, 2012 IEEE International Conference on Robotics and Automation.

[10]  Robert J. Peckham,et al.  Digital Terrain Modelling: Development and Applications in a Policy Support Environment (Lecture Notes in Geoinformation and Cartography) , 2007 .

[11]  Václav Hlavác,et al.  Adaptive traversability of partially occluded obstacles , 2015, 2015 IEEE International Conference on Robotics and Automation (ICRA).

[12]  Tomás Svoboda,et al.  TRADR Project: Long-Term Human-Robot Teaming for Robot Assisted Disaster Response , 2015, KI - Künstliche Intelligenz.

[13]  Joseph O'Rourke,et al.  Handbook of Discrete and Computational Geometry, Second Edition , 1997 .

[14]  Danica Kragic,et al.  Enhancing visual perception of shape through tactile glances , 2013, 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[15]  Marcos P. Gerardo-Castro,et al.  Laser-Radar Data Fusion with Gaussian Process Implicit Surfaces , 2013, FSR.

[16]  Mohammed El-Beltagy,et al.  Gaussian Processes for Model Fusion , 2001, ICANN.

[17]  Helge J. Ritter,et al.  A Probabilistic Approach to Tactile Shape Reconstruction , 2011, IEEE Transactions on Robotics.

[18]  Hans-Peter Seidel,et al.  Multi-level partition of unity implicits , 2003, ACM Trans. Graph..

[19]  Neil D. Lawrence,et al.  Fast Sparse Gaussian Process Methods: The Informative Vector Machine , 2002, NIPS.

[20]  A. Heyden,et al.  Reconstructing open surfaces from unorganized data points , 2004, CVPR 2004.

[21]  M. Ernst,et al.  Optimal integration of shape information from vision and touch , 2007, Experimental Brain Research.

[22]  Geoffrey A. Hollinger,et al.  Active planning for underwater inspection and the benefit of adaptivity , 2012, Int. J. Robotics Res..

[23]  Václav Hlavác,et al.  Adaptive Traversability of unknown complex terrain with obstacles for mobile robots , 2014, 2014 IEEE International Conference on Robotics and Automation (ICRA).

[24]  Martial Hebert,et al.  Natural terrain classification using three‐dimensional ladar data for ground robot mobility , 2006, J. Field Robotics.

[25]  Marc Toussaint,et al.  Gaussian process implicit surfaces for shape estimation and grasping , 2011, 2011 IEEE International Conference on Robotics and Automation.

[26]  W. Fisher,et al.  Hybrid Position/Force Control: A Correct Formulation , 1992 .

[27]  Fabio Tozeto Ramos,et al.  A Sparse Covariance Function for Exact Gaussian Process Inference in Large Datasets , 2009, IJCAI.

[28]  Shrihari Vasudevan,et al.  Data fusion with Gaussian processes , 2012, Robotics Auton. Syst..

[29]  Bernhard Schölkopf,et al.  Support Vector Machines for 3D Shape Processing , 2005, Comput. Graph. Forum.

[30]  Hugh F. Durrant-Whyte,et al.  Gaussian Process modeling of large scale terrain , 2009, 2009 IEEE International Conference on Robotics and Automation.

[31]  James F. Blinn,et al.  A Generalization of Algebraic Surface Drawing , 1982, TOGS.

[32]  Jonghyuk Kim,et al.  GPmap: A Unified Framework for Robotic Mapping Based on Sparse Gaussian Processes , 2013, FSR.

[33]  Gert Vegter,et al.  In handbook of discrete and computational geometry , 1997 .

[34]  Andrew Fitzgibbon,et al.  Gaussian Process Implicit Surfaces , 2006 .

[35]  R. Peckham,et al.  Digital Terrain Modelling , 2007 .

[36]  James F. O'Brien,et al.  Variational Implicit Surfaces , 1999 .