{
  "id": "2011.01758",
  "version": "v1",
  "published": "2020-11-03T15:00:36.000Z",
  "updated": "2020-11-03T15:00:36.000Z",
  "title": "Representation Matters: Improving Perception and Exploration for Robotics",
  "authors": [
    "Markus Wulfmeier",
    "Arunkumar Byravan",
    "Tim Hertweck",
    "Irina Higgins",
    "Ankush Gupta",
    "Tejas Kulkarni",
    "Malcolm Reynolds",
    "Denis Teplyashin",
    "Roland Hafner",
    "Thomas Lampe",
    "Martin Riedmiller"
  ],
  "categories": [
    "cs.LG",
    "cs.AI",
    "cs.RO",
    "stat.ML"
  ],
  "abstract": "Projecting high-dimensional environment observations into lower-dimensional structured representations can considerably improve data-efficiency for reinforcement learning in domains with limited data such as robotics. Can a single generally useful representation be found? In order to answer this question, it is important to understand how the representation will be used by the agent and what properties such a 'good' representation should have. In this paper we systematically evaluate a number of common learnt and hand-engineered representations in the context of three robotics tasks: lifting, stacking and pushing of 3D blocks. The representations are evaluated in two use-cases: as input to the agent, or as a source of auxiliary tasks. Furthermore, the value of each representation is evaluated in terms of three properties: dimensionality, observability and disentanglement. We can significantly improve performance in both use-cases and demonstrate that some representations can perform commensurate to simulator states as agent inputs. Finally, our results challenge common intuitions by demonstrating that: 1) dimensionality strongly matters for task generation, but is negligible for inputs, 2) observability of task-relevant aspects mostly affects the input representation use-case, and 3) disentanglement leads to better auxiliary tasks, but has only limited benefits for input representations. This work serves as a step towards a more systematic understanding of what makes a 'good' representation for control in robotics, enabling practitioners to make more informed choices for developing new learned or hand-engineered representations.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2020-11-03T15:00:36.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "representation matters",
      "improving perception",
      "exploration",
      "results challenge common intuitions",
      "hand-engineered representations"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}