{
  "id": "1901.08479",
  "version": "v1",
  "published": "2019-01-24T16:13:24.000Z",
  "updated": "2019-01-24T16:13:24.000Z",
  "title": "On the Transformation of Latent Space in Autoencoders",
  "authors": [
    "Jaehoon Cha",
    "Kyeong Soo Kim",
    "Sanghyuk Lee"
  ],
  "comment": "9 pages and 9 figures. The paper has been submitted to ICML (The International Conference on Machine Learning) 2019",
  "categories": [
    "cs.LG",
    "stat.ML"
  ],
  "abstract": "Noting the importance of the latent variables in inference and learning, we propose a novel framework for autoencoders based on the homeomorphic transformation of latent variables --- which could reduce the distance between vectors in the transformed space, while preserving the topological properties of the original space --- and investigate the effect of the transformation in both learning generative models and denoising corrupted data. The results of our experiments show that the proposed model can work as both a generative model and a denoising model with improved performance due to the transformation compared to conventional variational and denoising autoencoders.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2019-01-24T16:13:24.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "latent space",
      "autoencoders",
      "latent variables",
      "original space",
      "conventional variational"
    ],
    "tags": [
      "conference paper"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 9,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}