{
  "id": "1612.08185",
  "version": "v1",
  "published": "2016-12-24T14:20:05.000Z",
  "updated": "2016-12-24T14:20:05.000Z",
  "title": "Deep Probabilistic Modeling of Natural Images using a Pyramid Decomposition",
  "authors": [
    "Alexander Kolesnikov",
    "Christoph H. Lampert"
  ],
  "categories": [
    "cs.CV"
  ],
  "abstract": "We introduce a new technique for probabilistic modeling of natural images that combines the advantages of classic multi-scale and modern deep learning models. By explicitly representing natural images at different scales we derive a model that can capture high level image structure in a computationally efficient way. We show experimentally that our model achieves new state-of-the-art image modeling performance on the CIFAR-10 dataset and at the same time is much faster than competitive models. We also evaluate the proposed technique on a human faces dataset and demonstrate the potential of our model to generate nearly photorealistic face samples.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2016-12-24T14:20:05.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "natural images",
      "deep probabilistic modeling",
      "pyramid decomposition",
      "capture high level image structure",
      "state-of-the-art image modeling performance"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}