{
  "id": "1908.02893",
  "version": "v1",
  "published": "2019-08-08T01:00:11.000Z",
  "updated": "2019-08-08T01:00:11.000Z",
  "title": "EdgeNet: Semantic Scene Completion from RGB-D images",
  "authors": [
    "Aloisio Dourado",
    "Teofilo Emidio de Campos",
    "Hansung Kim",
    "Adrian Hilton"
  ],
  "comment": "10 pages, 5 figures",
  "categories": [
    "cs.CV"
  ],
  "abstract": "Semantic scene completion is the task of predicting a complete 3D representation of volumetric occupancy with corresponding semantic labels for a scene from a single point of view. Previous works on Semantic Scene Completion from RGB-D data used either only depth or depth with colour by projecting the 2D image into the 3D volume resulting in a sparse data representation. In this work, we present a new strategy to encode colour information in 3D space using edge detection and flipped truncated signed distance. We also present EdgeNet, a new end-to-end neural network architecture capable of handling features generated from the fusion of depth and edge information. Experimental results show improvement of 6.9% over the state-of-the-art result on real data, for end-to-end approaches.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2019-08-08T01:00:11.000Z"
    }
  ],
  "analyses": {
    "subjects": [
      "I.4.6",
      "I.4.8"
    ],
    "keywords": [
      "semantic scene completion",
      "rgb-d images",
      "encode colour information",
      "end-to-end neural network architecture capable",
      "complete 3d representation"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 10,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}