Unifying Geometry, Semantics, and Data Association in SLAM

Traditional approaches for simultaneous localization and mapping (SLAM) rely on geometric features such as points, lines, and planes to infer the environment structure. They make hard decisions about the (data) association between observed features and mapped landmarks to update the environment model. Our work makes two contributions to the state of the art in SLAM. First, it generalizes the purely geometric model by introducing semantically meaningful objects, represented as structured models of mid-level part features. Second, instead of making hard, potentially wrong associations between semantic features and objects, it shows that SLAM inference can be performed efficiently with probabilistic data association via matrix permanent computations. The approach not only allows building meaningful maps (containing doors, chairs, cars, etc.) but also offers significant advantages in ambiguous environments. Finally, one can go beyond SLAM and formulate high-level robot missions in terms of the objects on the map. Our work proposes algorithms for motion planning under temporal logic constraints in probabilistic semantic maps.

 

Related Publications

  • Probabilistic Data Association for Semantic SLAM
    S. Bowman, N. Atanasov, K. Daniilidis and G. Pappas
    IEEE Int. Conf. on Robotics and Automation (ICRA), 2017.
    [bib] [pdf] [doi]
  •   @inproceedings{Bowman_SemanticSLAM_ICRA17,
        author = {Sean Bowman and Nikolay Atanasov and Kostas Daniilidis and George Pappas},
        title = {Probabilistic Data Association for Semantic SLAM},
        booktitle = {IEEE Int. Conf. on Robotics and Automation (ICRA)},
        year = {2017},
        doi = {http://www.doi.org/10.1109/ICRA.2017.7989203}
      }
      
  • A Unifying View of Geometry, Semantics, and Data Association in SLAM
    N. Atanasov, S. Bowman, K. Daniilidis and G. Pappas
    International Joint Conference on Artificial Intelligence (IJCAI), 2018.
    [bib] [pdf] [doi]
  •   @inproceedings{Atanasov_SemanticSLAM_IJCAI18,
        author = {N. Atanasov and S. Bowman and K. Daniilidis and G. Pappas},
        title = {A Unifying View of Geometry, Semantics, and Data Association in SLAM},
        booktitle = {International Joint Conference on Artificial Intelligence (IJCAI)},
        year = {2018},
        doi = {http://www.doi.org/10.24963/ijcai.2018/722}
      }
      
  • Localization from Semantic Observations via the Matrix Permanent
    N. Atanasov, M. Zhu, K. Daniilidis and G. Pappas
    The International Journal of Robotics Research (IJRR), Vol. 35, pp. 73-99, 2015.
    [bib] [pdf] [doi]
  •   @article{Atanasov_SemanticLocalization_IJRR15,
        author = {N. Atanasov and M. Zhu and K. Daniilidis and G. Pappas},
        title = {Localization from Semantic Observations via the Matrix Permanent},
        journal = {The International Journal of Robotics Research (IJRR)},
        year = {2015},
        volume = {35},
        pages = {73-99},
        doi = {http://www.doi.org/10.1177/0278364915596589}
      }
      
  • Semantic Localization via the Matrix Permanent
    N. Atanasov, M. Zhu, K. Daniilidis and G. Pappas
    Robotics: Science and Systems (RSS), 2014.
    [bib] [pdf] [doi]
  •   @inproceedings{Atanasov_SemanticLocalization_RSS14,
        author = {N. Atanasov and M. Zhu and K. Daniilidis and G. Pappas},
        title = {Semantic Localization via the Matrix Permanent},
        booktitle = {Robotics: Science and Systems (RSS)},
        year = {2014},
        doi = {http://www.doi.org/10.15607/RSS.2014.X.043}
      }
      
  • Optimal Temporal Logic Planning in Probabilistic Semantic Maps
    J. Fu, N. Atanasov, U. Topcu and G. Pappas
    IEEE Int. Conf. on Robotics and Automation (ICRA), 2016.
    [bib] [pdf] [doi]
  •   @inproceedings{Fu_SemanticLTLPlanning_ICRA16,
        author = {Jie Fu and Nikolay Atanasov and Ufuk Topcu and George Pappas},
        title = {Optimal Temporal Logic Planning in Probabilistic Semantic Maps},
        booktitle = {IEEE Int. Conf. on Robotics and Automation (ICRA)},
        year = {2016},
        doi = {http://www.doi.org/10.1109/ICRA.2016.7487554}
      }