{
  "id": "2303.13825",
  "version": "v1",
  "published": "2023-03-24T06:19:19.000Z",
  "updated": "2023-03-24T06:19:19.000Z",
  "title": "HandNeRF: Neural Radiance Fields for Animatable Interacting Hands",
  "authors": [
    "Zhiyang Guo",
    "Wengang Zhou",
    "Min Wang",
    "Li Li",
    "Houqiang Li"
  ],
  "comment": "CVPR 2023",
  "categories": [
    "cs.CV"
  ],
  "abstract": "We propose a novel framework to reconstruct accurate appearance and geometry with neural radiance fields (NeRF) for interacting hands, enabling the rendering of photo-realistic images and videos for gesture animation from arbitrary views. Given multi-view images of a single hand or interacting hands, an off-the-shelf skeleton estimator is first employed to parameterize the hand poses. Then we design a pose-driven deformation field to establish correspondence from those different poses to a shared canonical space, where a pose-disentangled NeRF for one hand is optimized. Such unified modeling efficiently complements the geometry and texture cues in rarely-observed areas for both hands. Meanwhile, we further leverage the pose priors to generate pseudo depth maps as guidance for occlusion-aware density learning. Moreover, a neural feature distillation method is proposed to achieve cross-domain alignment for color optimization. We conduct extensive experiments to verify the merits of our proposed HandNeRF and report a series of state-of-the-art results both qualitatively and quantitatively on the large-scale InterHand2.6M dataset.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2023-03-24T06:19:19.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "neural radiance fields",
      "animatable interacting hands",
      "generate pseudo depth maps",
      "neural feature distillation method",
      "achieve cross-domain alignment"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}