multi-view-reconstruction 开源项目推荐（18 个）

autonomousvision /sdfstudio

表面重建的统一框架

GAP-LAB-CUHK-SZ /gaustudio

A Modular Framework for 3D Gaussian Splatting and Beyond

One-2-3-45 /One-2-3-45

[NeurIPS 2023]官方代码“One-2-3-45：45秒内将任何单个图像转换为3D网格，无需按形状优化”

autonomousvision /monosdf

[NeurIPS'22] MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction

yiqun-wang /PET-NeuS

PET-NeuS: Positional Encoding Tri-Planes for Neural Surfaces (CVPR 2023)

yiqun-wang /HFS

HF-NeuS: Improved Surface Reconstruction Using High-Frequency Details (NeurIPS 2022)

yulunwu0108 /NeuSurf

[AAAI'24] NeuSurf: On-Surface Priors for Neural Surface Reconstruction from Sparse Input Views

xucao-42 /mvas

[CVPR 2023] Multi-View Azimuth Stereo via Tangent Space Consistency

yiqun-wang /Factored-NeuS

Factored-NeuS: Reconstructing Surfaces, Illumination, and Materials of Possibly Glossy Objects (CVPR 2025)

LordLiang /GA-Sketching

(PG2023/CGF) This is the official PyTorch implementation of PG2023/CGF paper: GA-Sketching: Shape Modeling from Multi-View Sketching with Geometry-Aligned Deep Implicit Functions

YoujiaZhang /USD

[IJCAI 2025] Optimized View and Geometry Distillation from Multi-view Diffuser

zhijianglu /HF-Recon

High-Fidelity 3D Reconstruction via Graph Optimization. A Multi-Frame Registration and reconstruction algorithm.

hummat /sdfstudio

Actively maintained fork of SDFStudio: Neural implicit surface reconstruction with modern PyTorch, RTX 40XX support, PBR exports, and bug fixes

bwittmann /RecMVCNN

A lightweight neural network for multi-view classification and reconstruction.

timohaubner /multi_view_vdms_for_mesh_animation

Research code for evaluating multi-view video diffusion for humanoid mesh animation

saiga006 /Panoptic-MapAnything

A unified, feed-forward model integrating the MapAnything 3D reconstruction backbone with Mask2Former for multi-view consistent panoptic segmentation. This approach provides feature-rich and metric accurate semantic maps for indoor robot navigation while eliminating the need for posed inputs and costly test-time optimization.

SandeepaInduwaraSamaranayake /SwinVox

A novel hybrid architecture combining CNNs and Swin Transformers, to address limitations in 3D voxel reconstruction.

PetropoulakisPanagiotis /Multi-View-RGB-based-Recognition-and-Reconstruction

Multi-View RGB-based Recognition and Reconstruction in PyTorch