{
  "id": "2006.04569",
  "version": "v1",
  "published": "2020-06-08T13:20:33.000Z",
  "updated": "2020-06-08T13:20:33.000Z",
  "title": "Person Re-identification in the 3D Space",
  "authors": [
    "Zhedong Zheng",
    "Yi Yang"
  ],
  "comment": "The code is available at https://github.com/layumi/person-reid-3d",
  "categories": [
    "cs.CV"
  ],
  "abstract": "People live in a 3D world. However, existing works on person re-identification (re-id) mostly consider the representation learning in a 2D space, intrinsically limiting the understanding of people. In this work, we address this limitation by exploring the prior knowledge of the 3D body structure. Specifically, we project 2D images to a 3D space and introduce a novel Omni-scale Graph Network (OG-Net) to learn the representation from sparse 3D points. With the help of 3D geometry information, we can learn a new type of deep re-id feature free from noisy variants, such as scale and viewpoint. To our knowledge, we are among the first attempts to conduct person re-identification in the 3D space. Extensive experiments show that the proposed method achieves competitive results on three popular large-scale person re-id datasets, and has good scalability to unseen datasets.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2020-06-08T13:20:33.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "3d space",
      "popular large-scale person re-id datasets",
      "novel omni-scale graph network",
      "deep re-id feature free",
      "3d body structure"
    ],
    "tags": [
      "github project"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}