{
  "id": "1805.08957",
  "version": "v1",
  "published": "2018-05-23T04:26:50.000Z",
  "updated": "2018-05-23T04:26:50.000Z",
  "title": "Semi-Supervised Learning with GANs: Revisiting Manifold Regularization",
  "authors": [
    "Bruno Lecouat",
    "Chuan-Sheng Foo",
    "Houssam Zenati",
    "Vijay R. Chandrasekhar"
  ],
  "comment": "Accepted paper",
  "journal": "Workshop track - ICLR 2018",
  "categories": [
    "cs.LG",
    "stat.ML"
  ],
  "abstract": "GANS are powerful generative models that are able to model the manifold of natural images. We leverage this property to perform manifold regularization by approximating the Laplacian norm using a Monte Carlo approximation that is easily computed with the GAN. When incorporated into the feature-matching GAN of Improved GAN, we achieve state-of-the-art results for GAN-based semi-supervised learning on the CIFAR-10 dataset, with a method that is significantly easier to implement than competing methods.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2018-05-23T04:26:50.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "revisiting manifold regularization",
      "semi-supervised learning",
      "achieve state-of-the-art results",
      "perform manifold regularization",
      "monte carlo approximation"
    ],
    "tags": [
      "journal article"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}