Learning Visibility Field for Detailed 3D Human Reconstruction and Relighting

• URL: http://arxiv.org/abs/2304.11900v1
• Date: Mon, 24 Apr 2023 08:19:03 GMT
• Title: Learning Visibility Field for Detailed 3D Human Reconstruction and Relighting
• Authors: Ruichen Zheng and Peng Li and Haoqian Wang and Tao Yu
• Abstract summary: We propose a novel sparse-view 3d human reconstruction framework that closely incorporates the occupancy field and albedo field with an additional visibility field. Results and experiments demonstrate the effectiveness of the proposed method, as it surpasses state-of-the-art in terms of reconstruction accuracy.
• Score: 19.888346124475042
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Detailed 3D reconstruction and photo-realistic relighting of digital humans are essential for various applications. To this end, we propose a novel sparse-view 3d human reconstruction framework that closely incorporates the occupancy field and albedo field with an additional visibility field--it not only resolves occlusion ambiguity in multiview feature aggregation, but can also be used to evaluate light attenuation for self-shadowed relighting. To enhance its training viability and efficiency, we discretize visibility onto a fixed set of sample directions and supply it with coupled geometric 3D depth feature and local 2D image feature. We further propose a novel rendering-inspired loss, namely TransferLoss, to implicitly enforce the alignment between visibility and occupancy field, enabling end-to-end joint training. Results and extensive experiments demonstrate the effectiveness of the proposed method, as it surpasses state-of-the-art in terms of reconstruction accuracy while achieving comparably accurate relighting to ray-traced ground truth.

Related papers

• Depth Reconstruction with Neural Signed Distance Fields in Structured Light Systems [15.603880588503355]
  We introduce a novel depth estimation technique for multi-frame structured light setups using neural implicit representations of 3D space. Our approach employs a neural signed distance field (SDF), trained through self-supervised differentiable rendering.
  arXiv  Detail & Related papers  (2024-05-20T13:24:35Z)
• Relightable 3D Gaussians: Realistic Point Cloud Relighting with BRDF Decomposition and Ray Tracing [21.498078188364566]
  We present a novel differentiable point-based rendering framework to achieve photo-realistic relighting. The proposed framework showcases the potential to revolutionize the mesh-based graphics pipeline with a point-based pipeline enabling editing, tracing, and relighting.
  arXiv  Detail & Related papers  (2023-11-27T18:07:58Z)
• Robust Geometry-Preserving Depth Estimation Using Differentiable Rendering [93.94371335579321]
  We propose a learning framework that trains models to predict geometry-preserving depth without requiring extra data or annotations. Comprehensive experiments underscore our framework's superior generalization capabilities. Our innovative loss functions empower the model to autonomously recover domain-specific scale-and-shift coefficients.
  arXiv  Detail & Related papers  (2023-09-18T12:36:39Z)
• Making the Invisible Visible: Toward High-Quality Terahertz Tomographic Imaging via Physics-Guided Restoration [24.045067900801072]
  Terahertz (THz) tomographic imaging has recently attracted significant attention thanks to its non-invasive, non-destructive, non-ionizing, material-classification, and ultra-fast nature for object exploration and inspection. We propose a novel multi-view Subspace-guided Restoration Network (SARNet) that fuses multi-viewAttention and multi-spectral features of THz images for effective image restoration and 3D tomographic reconstruction.
  arXiv  Detail & Related papers  (2023-04-28T15:05:46Z)
• Neural Light Field Estimation for Street Scenes with Differentiable Virtual Object Insertion [129.52943959497665]
  Existing works on outdoor lighting estimation typically simplify the scene lighting into an environment map. We propose a neural approach that estimates the 5D HDR light field from a single image. We show the benefits of our AR object insertion in an autonomous driving application.
  arXiv  Detail & Related papers  (2022-08-19T17:59:16Z)
• Neural 3D Reconstruction in the Wild [86.6264706256377]
  We introduce a new method that enables efficient and accurate surface reconstruction from Internet photo collections. We present a new benchmark and protocol for evaluating reconstruction performance on such in-the-wild scenes.
  arXiv  Detail & Related papers  (2022-05-25T17:59:53Z)
• Unsupervised Discovery and Composition of Object Light Fields [57.198174741004095]
  We propose to represent objects in an object-centric, compositional scene representation as light fields. We propose a novel light field compositor module that enables reconstructing the global light field from a set of object-centric light fields.
  arXiv  Detail & Related papers  (2022-05-08T17:50:35Z)
• Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing [41.34640834483265]
  We present PHORHUM, a novel, end-to-end trainable, deep neural network methodology for photorealistic 3D human reconstruction given just a monocular RGB image. Our pixel-aligned method estimates detailed 3D geometry and, for the first time, the unshaded surface color together with the scene illumination.
  arXiv  Detail & Related papers  (2022-04-19T14:06:16Z)
• Learning Indoor Inverse Rendering with 3D Spatially-Varying Lighting [149.1673041605155]
  We address the problem of jointly estimating albedo, normals, depth and 3D spatially-varying lighting from a single image. Most existing methods formulate the task as image-to-image translation, ignoring the 3D properties of the scene. We propose a unified, learning-based inverse framework that formulates 3D spatially-varying lighting.
  arXiv  Detail & Related papers  (2021-09-13T15:29:03Z)
• Deep 3D Capture: Geometry and Reflectance from Sparse Multi-View Images [59.906948203578544]
  We introduce a novel learning-based method to reconstruct the high-quality geometry and complex, spatially-varying BRDF of an arbitrary object. We first estimate per-view depth maps using a deep multi-view stereo network. These depth maps are used to coarsely align the different views. We propose a novel multi-view reflectance estimation network architecture.
  arXiv  Detail & Related papers  (2020-03-27T21:28:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.