Pratik Agarwal

PhD Thesis

Robust Graph-Based Localization and Mapping, PhD thesis.
Preprint FreiDok BibTeX

@PHDTHESIS{agarwal15phd,
  author = {Pratik Agarwal},
  title = {Robust Graph-Based Localization and Mapping},
  school = {University of Freiburg},
  year = {2015},
  address = {Freiburg, Germany}
}

Patents

• Jiajun Zhu and Pratik Agarwal
  Methods and Systems for Object Detection using Multiple Sensors.
  US Patent number 9,098,753.
  Issued on August 4, 2015.
  
• Pratik Agarwal and Jiajun Zhu
  Estimating multi-vehicle motion characteristics by finding stable reference points.
  US Patent number 8,886,387.
  Issued: November 11, 2014
  
• Pratik Agarwal, Jiajun Zhu and Dmitri Dolgov
  Methods and devices for determining movements of an object in an environment.
  US Patent number 8,989,944.
  Issued: March 24, 2015
  

Journal Papers

• Pratik Agarwal, Wolfram Burgard and Cyrill Stachniss
  A Survey of Geodetic Approaches to Mapping and the Relationship to Graph-Based SLAM
  Robotics and Automation Magazine, September 2014
  Preprint BibTeX

  @article{agarwal14ram,
    author = {Pratik Agarwal and Wolfram Burgard and Cyrill Stachniss},
    title = {A Survey of Geodetic Approaches to Mapping and the Relationship to Graph-Based SLAM},
    journal = {Robotics and Automation Magazine},
    year = 2014,
    month = Sep,
    doi = {10.1109/MRA.2014.2322282},
    url = {http://ais.informatik.uni-freiburg.de/publications/papers/agarwal-geodetic.pdf}
  }
  

  Abstract

• Edwin Olson and Pratik Agarwal
  Inference on networks of mixtures for robust robot mapping
  International Journal of Robotics Research, 2013
  Preprint Online first BibTeX

  @article{olson2013ijrr,
      AUTHOR     = {Edwin Olson and Pratik Agarwal},
      TITLE      = {Inference on networks of mixtures for robust robot mapping},
      JOURNAL    = {International Journal of Robotics Research},
      VOLUME     = {32},
      NUMBER     = {7},
      YEAR       = {2013},
      MONTH      = {July},
      PAGES      = {826-840},
  }

  Abstract

  The central challenge in robotic mapping is obtaining reliable data associations (or "loop
  closures"): state-of-the-art inference algorithms can fail catastrophically if even one erroneous
  loop closure is incorporated into the map. Consequently, much work has been done to push
  error rates closer to zero. However, a long-lived or multi-robot system will still encounter errors,
  leading to system failure.
  We propose a fundamentally different approach: allow richer error models that allow the
  probability of a failure to be explicitly modeled. In other words, rather than characterizing loop
  closures as being "right" or "wrong", we propose characterizing the error of those loop closures
  in a more expressive manner that can account for their non-Gaussian behavior. Our approach
  leads to an fully-integrated Bayesian framework for dealing with error-prone data. Unlike earlier
  multiple-hypothesis approaches, our approach avoids exponential memory complexity and is fast
  enough for real-time performance.
  We show that the proposed method not only allows loop closing errors to be automatically
  identified, but also that in extreme cases, the "front-end" loop-validation systems can be unnecessary.
  We demonstrate our system both on standard benchmarks and on the real-world
  datasets that motivated this work.
  

Conference Papers

• Pratik Agarwal, Wolfram Burgard and Luciano Spinello
  Metric Localization using Google Street View
  Proceedings of the International Conference on Intelligent Robots and Systems (IROS), Hamburg, Germany, 2015
  Download BibTeX

    @inproceedings{agarwal15iros,
      TITLE      = {Metric Localization using Google Street View},
      AUTHOR     = {Pratik Agarwal and Wolfram Burgard and Luciano Spinello},
      BOOKTITLE  = {Proceedings of the {IEEE/RSJ} International Conference on Intelligent
                   Robots and Systems {(IROS)}},
      YEAR       = {2015},
      MONTH      = {October},
      KEYWORDS   = {localization, panoramas, street-view}
  }
  

• Bahram Behzadian, Pratik Agarwal, Wolfram Burgard and Gian Diego Tipaldi
  Montel Carlo Localization in Hand-Drawn Maps
  Proceedings of the International Conference on Intelligent Robots and Systems (IROS), Hamburg, Germany, 2015
  Download BibTeX

    @inproceedings{bahram15iros,
      TITLE      = {Monte Carlo Localization in Hand-Drawn Maps},
      AUTHOR     = {Bharam Behzadian and Pratik Agarwal and Wolfram Burgard and Gian Diego Tipaldi},
      BOOKTITLE  = {Proceedings of the {IEEE/RSJ} International Conference on Intelligent
                   Robots and Systems {(IROS)}},
      YEAR       = {2015},
      MONTH      = {October},
      KEYWORDS   = {localization, sketch-maps}
  }
  

• Andreas Wachaja, Pratik Agarwal, Mathias Zink, Miguel Reyes Adame, Knut Möller and Wolfram Burgard
  Naviating Blind People with a Smart Walker
  Proceedings of the International Conference on Intelligent Robots and Systems (IROS), Hamburg, Germany, 2015
  Download BibTeX

    @inproceedings{wachaja15iros,
      TITLE      = {Navigating Blind People with a Smart Walker},
      AUTHOR     = {Andreas Wachaja, Pratik Agarwal, Mathias Zink, Miguel Reyes Adame, Knut M\"oller and Wolfram Burgad},
      BOOKTITLE  = {Proceedings of the {IEEE/RSJ} International Conference on Intelligent
                   Robots and Systems {(IROS)}},
      YEAR       = {2015},
      MONTH      = {October},
      KEYWORDS   = {walker, blind, navigation}
  }
  

• Pratik Agarwal, Giorgio Grisetti, Gian Diego Tipaldi, Luciano Spinello, Wolfram Burgard and Cyrill Stachniss
  Experimental Analysis of Dynamic Covariance Scaling for Robust Map Optimization Under Bad Initial Estimates
  Proceedings of the International Conference on Robots and Automation (ICRA), Hong Kong, 2014.
  Download Presentation BibTeX

  @inproceedings{agarwal2014aicra,
      TITLE      = {Experimental Analysis of Dynamic Covariance Scaling for Robust Map Optimization Under Bad Initial Estimates},
      AUTHOR     = {Pratik Agarwal and Giorgio Grisetti and Gian Diego Tipaldi and Luciano Spinello and Wolfram Burgard and Cyrill Stachniss},
      BOOKTITLE  = { Proceedings of the {IEEE} International Conference on Robotics and
                   Automation ({ICRA}) },
      YEAR       = {2014},
      KEYWORDS   = {SLAM, Robust optimization}
  }

• Pratik Agarwal, Wolfram Burgard and Cyrill Stachniss
  Helmert's and Bowie's Geodetic Mapping Methods and Their Relationship to Graph-Based SLAM
  Proceedings of the International Conference on Robots and Automation (ICRA), Hong Kong, 2014.
  Download Presentation BibTeX

  @inproceedings{agarwal2014bicra,
      TITLE      = {Helmert's and Bowie's Geodetic Mapping Methods and Their Relationship to Graph-Based SLAM},
      AUTHOR     = {Pratik Agarwal and Wolfram Burgard and Cyrill Stachniss},
      BOOKTITLE  = { Proceedings of the {IEEE} International Conference on Robotics and
                   Automation ({ICRA}) },
      YEAR       = {2014},
      KEYWORDS   = {SLAM, Survey, Geodetic}
  }

• Pratik Agarwal, Gian Diego Tipaldi, Luciano Spinello, Cyrill Stachniss and Wolfram Burgard
  Robust Map Optimization using Dynamic Covariance Scaling
  Proceedings of the International Conference on Robots and Automation (ICRA), Karlsruhe, Germany, 2013
  Download BibTeX

  @inproceedings{agarwal2013aicra,
      TITLE      = {Robust Map Optimization using Dynamic Covariance Scaling},
      AUTHOR     = {Pratik Agarwal and Gian Diego Tipaldi and Luciano Spinello and Cyrill Stachniss and Wolfram Burgard },
      BOOKTITLE  = { Proceedings of the {IEEE} International Conference on Robotics and
                   Automation ({ICRA}) },
      YEAR       = {2013},
      MONTH      = {May},
      KEYWORDS   = {SLAM, Robust optimization}
  }

  Abstract

  Abstract—Developing the perfect SLAM front-end that produces
  graphs which are free of outliers is generally impossible due
  to perceptual aliasing. Therefore, optimization back-ends need to
  be able to deal with outliers resulting from an imperfect frontend.
  In this paper, we introduce dynamic covariance scaling, a
  novel approach for effective optimization of constraint networks
  under the presence of outliers. The key idea is to use a
  robust function that generalizes classical gating and dynamically
  rejects outliers without compromising convergence speed. We
  implemented and thoroughly evaluated our method on publicly
  available datasets. Compared to recently published state-of-theart
  methods, we obtain a substantial speed up without increasing
  the number of variables in the optimization process. Our method
  can be easily integrated in almost any SLAM back-end.
  

  Video
• Edwin Olson and Pratik Agarwal
  Inference on networks of mixtures for robust robot mapping
  Proceedings of Robotics: Science and Systems (RSS), Sydney, Australia, 2012
  Download BibTeX

  @inproceedings{olson2012rss,
      AUTHOR     = {Edwin Olson and Pratik Agarwal},
      TITLE      = {Inference on networks of mixtures for robust robot mapping},
      BOOKTITLE  = {Proceedings of Robotics: Science and Systems ({RSS})},
      YEAR       = {2012},
      MONTH      = {July},
      ADDRESS    = {Sydney, Australia},
  }

  Abstract

     The central challenge in robotic mapping is obtaining
  reliable data associations (or "loop closures"): state-ofthe-
  art inference algorithms can fail catastrophically if even
  one erroneous loop closure is incorporated into the map.
  Consequently, much work has been done to push error rates
  closer to zero. However, a long-lived or multi-robot system will
  still encounter errors, leading to system failure.
  We propose a fundamentally different approach: allow richer
  error models that allow the probability of a failure to be
  explicitly modeled. In other words, we optimize the map while
  simultaneously determining which loop closures are correct
  from within a single integrated Bayesian framework. Unlike
  earlier multiple-hypothesis approaches, our approach avoids
  exponential memory complexity and is fast enough for realtime
  performance.
  We show that the proposed method not only allows loop
  closing errors to be automatically identified, but also that in
  extreme cases, the "front-end" loop-validation systems can be
  unnecessary. We demonstrate our system both on standard
  benchmarks and on the real-world datasets that motivated this
  work.
     

  City10000 CSW3500
• Pratik Agarwal and Edwin Olson
  Variable reordering strategies for SLAM
  Proceedings of the International Conference on Intelligent Robots and Systems (IROS), Vilamoura, Portugal, 2012
  Download BibTeX

    @inproceedings{agarwal2012iros,
      TITLE      = {Variable reordering strategies for SLAM},
      AUTHOR     = {Pratik Agarwal and Edwin Olson},
      BOOKTITLE  = {Proceedings of the {IEEE/RSJ} International Conference on Intelligent
                   Robots and Systems {(IROS)}},
      YEAR       = {2012},
      MONTH      = {October},
      KEYWORDS   = {SLAM, matrix factorization}
  }
  

  Abstract

  State of the art methods for state estimation and
  perception make use of least-squares optimization methods
  to perform efficient inference on noisy sensor data. Much of
  this efficiency is achieved by using sparse matrix factorization
  methods. The sparsity structure of the underlying matrix
  factorization which makes these optimization methods tractable
  is highly dependent on the choice of variable reordering; but
  there has been no systematic evaluation of reordering methods
  in the SLAM community.
  In this paper we evaluate the performance of various
  reordering techniques on benchmark SLAM data sets and
  provide definitive recommendations based on our results. We
  also compare these state of the art algorithms against our simple
  and easy to implement algorithm which achieves comparable
  performance. Finally, we provide empirical evidence that few
  gains remain with respect to variants of minimum degree
  ordering.
     

Preprints

• Pratik Agarwal, Wolfram Burgard and Luciano Spinello
  Metric Localization using Google Street View
  Download BibTeX

  @inproceedings{agarwal15arxiv,
      TITLE      = {Metric Localization using Google Street View},
      AUTHOR     = {Pratik Agarwal, Wolfram Burgard and Luciano Spinello},
      BOOKTITLE  = {arXiv preprint arXiv:1503.04287},
      YEAR       = {2015},
  }

Workshop & Extended Abstracts.

• Andreas Wachaja, Pratik Agarwal, Miguel Reyes Adame, Knut Möller and Wolfram Burgard
  A Navigation Aid for Blind People with Walking Disabilities
  IROS Workshop on Rehabilitation & Assistive Robotics
  Download BibTeX

  @inproceedings{wachaja2014iros,
      TITLE      = {A Navigation Aid for Blind People with Walking Disabilities}
      AUTHOR     = { Andreas Wachaja,  Pratik Agarwal,  Miguel Reyes Adame,  Knut M\"oller and Wolfram Burgard },
      BOOKTITLE  = { IROS Workshop on Rehabilitation \& Assistive Robotics, {IROS},
      YEAR       = {2014},
      MONTH      = {November},
      KEYWORDS   = {Rehabilitation, Robotics}
  }
  

• Miguel Reyes Adame, Andreas Lars Wachaja, Pratik Agarwal, Knut Möller and Wolfram Burgard
  Development of a Smart Walker with a Vibrating Belt for Assisting Visually Impaired.
  BMT - Biomedical Technology/ Biomedical Engineering
  Download BibTeX

  @inproceedings{miguel2014bmt,
      TITLE      = {Development of a Smart Walker with a Vibrating Belt for Assisting Visually Impaired.  }
      AUTHOR     = {Miguel Reyes Adame, Andreas Lars Wachaja, Pratik Agarwal, Knut M\"oeller and Wolfram Burgard},
      BOOKTITLE  = {BMT - Biomedical Technology/ Biomedical Engineering },
      YEAR       = {2014},
      KEYWORDS   = {Smart walker, navigation aid, blind}
  }
  

• Pratik Agarwal, Gian Diego Tipaldi, Luciano Spinello, Cyrill Stachniss and Wolfram Burgard
  Dynamic Covariance Scaling for Robust Robotic Mapping
  Workshop on robust and Multimodal Inference in Factor Graphs held at ICRA, Karlsruhe, Germany, 2013
  Download Presentation BibTeX

  @inproceedings{agarwal2013bicra,
      TITLE      = {Dynamic Covariance Scaling for Robust Robotic Mapping}
      AUTHOR     = {Pratik Agarwal and Gian Diego Tipaldi and Luciano Spinello and Cyrill Stachniss and Wolfram Burgard },
      BOOKTITLE  = {Workshop on robust and Multimodal Inference in Factor Graphs, {ICRA}},
      YEAR       = {2013},
      MONTH      = {May},
      KEYWORDS   = {SLAM, Robust optimization}
  }
  

  Abstract

  Developing the perfect SLAM front-end that produces
  graphs which are free of outliers is hard to achieve due
  to perceptual aliasing. Converging to the correct solution is
  challenging for non-linear error minimization SLAM techniques
  even in the absence of outliers, if the initial guess is far away
  from the correct solution. Therefore, optimization back-ends
  need to be resilient to outliers resulting from an imperfect
  front-end as well as be robust to bad initialization. In this
  paper, we present dynamic covariance scaling, a novel approach
  for effective optimization of constraint networks under the
  presence of outliers and bad initial guess. The key idea is
  to use a robust function that generalizes classical gating and
  down-weights outliers without compromising convergence speed.
  Compared to recently published state-of-the-art methods, we
  obtain a substantial speed-up without increasing overheads.