{
  "id": "1908.01301",
  "version": "v1",
  "published": "2019-08-04T09:37:24.000Z",
  "updated": "2019-08-04T09:37:24.000Z",
  "title": "Adversarial View-Consistent Learning for Monocular Depth Estimation",
  "authors": [
    "Yixuan Liu",
    "Yuwang Wang",
    "Shengjin Wang"
  ],
  "comment": "BMVC 2019 Spotlight",
  "categories": [
    "cs.CV",
    "cs.LG",
    "eess.IV"
  ],
  "abstract": "This paper addresses the problem of Monocular Depth Estimation (MDE). Existing approaches on MDE usually model it as a pixel-level regression problem, ignoring the underlying geometry property. We empirically find this may result in sub-optimal solution: while the predicted depth map presents small loss value in one specific view, it may exhibit large loss if viewed in different directions. In this paper, inspired by multi-view stereo (MVS), we propose an Adversarial View-Consistent Learning (AVCL) framework to force the estimated depth map to be all reasonable viewed from multiple views. To this end, we first design a differentiable depth map warping operation, which is end-to-end trainable, and then propose a pose generator to generate novel views for a given image in an adversarial manner. Collaborating with the differentiable depth map warping operation, the pose generator encourages the depth estimation network to learn from hard views, hence produce view-consistent depth maps . We evaluate our method on NYU Depth V2 dataset and the experimental results show promising performance gain upon state-of-the-art MDE approaches.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2019-08-04T09:37:24.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "monocular depth estimation",
      "adversarial view-consistent learning",
      "differentiable depth map warping operation",
      "nyu depth v2 dataset",
      "pose generator"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}