Surface light fields for 3D photography

By Daniel N. Wood, D. I. Azuma, Kenneth R. Aldinger, Brian Curless, T. Duchamp, David Salesin, Werner Stuetzle. International Conference on Computer Graphics and Interactive Techniques, 2000

559 citations recorded by the source corpus.

Subfields: Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/1cf2aef3a63465b99c38d9aaaaeb4c373bb00998

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

Citations in the corpus

Multiview Imaging and 3DTV (2007, methodology): “In [22], a camera attached to a spherical gantry arm is used to capture images roughly evenly over the sphere.”
Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction (2021, background): “Mesh-based representations [8, 54, 58, 63] can run in real-time but have a hard time with gradient-based optimization without template meshes provided.”
SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering (2023, background): “Such algorithms consider textured triangles or warp and blend captured images on the mesh surface to generate novel views [4, 12, 37].”
Example-based photometric stereo: shape reconstruction with general, varying BRDFs (2005, background): “To avoid such artifacts, shiny objects are commonly painted with a diffuse coating prior to laser scanning [6], [32]—we did not do this for the bottle, as we wanted to demonstrate these artifacts.”
Local light field fusion (2019, unknown): “mic viewing with mostly rotational and little translational viewer movement [Hedman et al. 2017; Hedman and Kopf 2018], but this strategy can only simulate Lambertian materials. Surface light fields [Wood et al. 2000] are able to render convincing view-dependent effects, but they require accurate ge”
Survey of image-based representations and compression techniques (2003, background): “that 3-D scene geometry can improve coding efficiency and rendering quality considerably [14], [88].”
PhySG: Inverse Rendering with Spherical Gaussians for Physics-based Material Editing and Relighting (2021, methodology): “Synthetic Kitty Synthetic Bear Synthetic Duck Synthetic Mouse SLF fish [50] Chips Corncho1 [35] # train/test (HxW) 100/100 (512x512) 100/100 (512x512) 100/100 (512x512) 100/100 (512x512) 419/106 (480x640) 1661/345 (480x640) ↓LPIPS ↑SSIM ↑PSNR ↓LPIPS ↑SSIM ↑PSNR ↓LPIPS ↑SSIM ↑PSNR ↓LPIPS ↑SSIM ↑PSNR ”
Unstructured lumigraph rendering (2001, unknown): “Table 1: Comparison of the algorithms lh96 [10], gor96 [5], deb96 [3], pul97 [13], deb98 [4], pigh98 [12], hei99 [7], wood00 [18], and ULR according to our desired goals.”
NeX: Real-time View Synthesis with Neural Basis Expansion (2021, methodology): “Our reparameterization of pixel into a combination of basis functions is closely related to light field factorization approaches in many areas, such as surface light field [48], precomputed radiance transfer [36, 35], BRDF estimations [11], light field and tensor display [46].”
Dense Depth Priors for Neural Radiance Fields from Sparse Input Views (2021, background): “A common paradigm for geometry-based view synthesis is to use a triangle mesh representation of scene geometry to reproject observed images into each novel viewpoint and combine them using either heuristic [1,9,27] or learned [14,22] blending algorithms.”
Synthesis of bidirectional texture functions on arbitrary surfaces (2002, background): “For surface light fields (e.g., [ 21 ]), the appearance of surfaces under fixed light sources but different view directions are captured and stored in a compact way for rendering.”
Neural Point-Based Graphics (2019, background): “processes determine the photometric properties of surface fragments and store them in the form of 2D parameterized maps, such as texture maps [4], bump maps [3], view-dependent textures [9], surface lightfields [48].”
Editing Conditional Radiance Fields (2021, background): “Prior work reasons over an explicit [11, 15, 22, 72] or discrete volumetric [20, 33, 36, 40, 44, 55, 64, 65, 67, 68, 80, 82] representation of the underlying geometry.”
A Comparison and Evaluation of Multi-View Stereo Reconstruction Algorithms (2006, methodology): “Similar data are available for surface lightfield studies [59, 60]; we have followed similar procedures for acquiring the images and models and for registering them to one another (although we add a step to automatically refine the alignment of the ground truth to the image sets based on minimizing ”