Paper Reading

Iterative Poisson Surface Reconstruction (iPSR) for Unoriented Points [PDF] [code]

迭代泊松重建和调整输入点的法线

SimpleRecon: 3D Reconstruction Without 3D Convolutions [PDF] [code]

对带位姿的图片估计深度来三维重建，cost volume中除了图片特征，还引入位姿和方向等。 allows for online real-time low-memory reconstruction

Deconstructing Self-Supervised Monocular Reconstruction: The Design Decisions that Matter [PDF]

提出修正的单目深度估计数据集并对之前的方法重新评估对比

NICE-SLAM: Neural Implicit Scalable Encoding for SLAM [PDF] [Project]

以渲染的深度图、颜色图来优化场景表示, 同时优化相机位姿。

Learning to Generate Realistic LiDAR Point Clouds [ECCV22]

竟然还有这种研究怎么生成lidar点云的文章。。 还有IROS22

Polarimetric Inverse Rendering for Transparent Shapes Reconstruction [code]

Neural 3D Reconstruction in the Wild [SIGGRAPH 2022]

隐式场重建的工作还有：UNISURF ICCV21, NeuS, VolSDF NIPS21

Mega-NeRF: Scalable Construction of Large-Scale NeRFs for Virtual Fly-Throughs CVPR22

NeRF的分块

BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale Scene Rendering ECCV22

Multi-View Reconstruction using Signed Ray Distance Functions [PDF]

ZeroMesh: Zero-shot Single-view 3D Mesh Reconstruction [PDF]

MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient Neural Field Rendering on Mobile Architectures

利用传统渲染管线快速渲染NeRF [Project Page]

同样地：Baking Neural Radiance Fields for Real-Time View Synthesis [ICCV21]

Fourier PlenOctrees for Dynamic Radiance Field Rendering in Real-time [CVPR22]

Urban Radiance Fields [CVPR22]

NPBG++: Accelerating Neural Point-Based Graphics [CVPR22]

Instant Neural Graphics Primitives

PlenOctrees for Real-time Rendering of Neural Radiance Fields

Stochastic Poisson Surface Reconstruction (https://arxiv.org/abs/2206.15236v2)