{
  "id": "2107.04481",
  "version": "v1",
  "published": "2021-07-09T15:12:55.000Z",
  "updated": "2021-07-09T15:12:55.000Z",
  "title": "Semantic and Geometric Unfolding of StyleGAN Latent Space",
  "authors": [
    "Mustafa Shukor",
    "Xu Yao",
    "Bharath Bhushan Damodaran",
    "Pierre Hellier"
  ],
  "comment": "16 pages",
  "categories": [
    "cs.CV"
  ],
  "abstract": "Generative adversarial networks (GANs) have proven to be surprisingly efficient for image editing by inverting and manipulating the latent code corresponding to a natural image. This property emerges from the disentangled nature of the latent space. In this paper, we identify two geometric limitations of such latent space: (a) euclidean distances differ from image perceptual distance, and (b) disentanglement is not optimal and facial attribute separation using linear model is a limiting hypothesis. We thus propose a new method to learn a proxy latent representation using normalizing flows to remedy these limitations, and show that this leads to a more efficient space for face image editing.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2021-07-09T15:12:55.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "stylegan latent space",
      "geometric unfolding",
      "proxy latent representation",
      "facial attribute separation",
      "image perceptual distance"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 16,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}