doublez0108/mvs

Multi-View Stereo based on deep learning. | Learning notes, codes and more.

MVS

Multi-View Stereo based on deep learning

📃论文列表

✨MVS系列核心创新点 & 缺陷

	简称/笔记	论文题目	出处(年份)	原文 代码	推荐值
1	MVSNet	MVSNet: Depth Inference for Unstructured Multi-view Stereo	ECCV 2018	paper code	★★★★★
2	R-MVSNet	Recurrent MVSNet for High-resolution Multi-view Stereo Depth Inference	CVPR 2019	paper code	★★★
3	Point-MVSNet	Point-based multi-view stereo network	ICCV 2019 oral	paper code	★★★★
4	P-MVSNet	P-MVSNet: Learning Patch-wise Matching Confidence Aggregation for Multi-View Stereo	ICCV 2019	paper	★
5	CVP-MVSNet	Cost Volume Pyramid Based Depth Inference for Multi-View Stereo	CVPR 2020 oral	paper code	★★★★★
6	RayNet	RayNet: Learning Volumetric 3D Reconstruction with Ray Potentials	CVPR 2018 oral	paper code	★★★☆
7	AttMVSNet	Attention-Aware Multi-View Stereo	CVPR 2020	paper	★☆
8	CasMVSNet	Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching	CVPR 2020 oral	paper code	★★★★★
9	PatchmatchNet	PatchmatchNet: Learned Multi-View Patchmatch Stereo	CVPR 2021 oral	paper code	★★★☆
10	MVSCRF	MVSCRF: Learning Multi-View Stereo With Conditional Random Fields	ICCV 2019	paper	★★☆
11	PVA-MVSNet	Pyramid Multi-view Stereo Net with Self-adaptive View Aggregation	ECCV 2020	paper code	★★★
12	FastMVSNet	Fast-MVSNet: Sparse-to-Dense Multi-View Stereo With Learned Propagation and Gauss-Newton Refinement	CVPR 2020	paper code	★★☆
13	UCSNet	Deep Stereo using Adaptive Thin Volume Representation with Uncertainty Awareness	CVPR 2020 oral	paper code	★★★★☆
14	MVSNet++	MVSNet++: Learning Depth-Based Attention Pyramid Features for Multi-View Stereo	TIP 2020	paper	★☆
15	SurfaceNet+	SurfaceNet+: An End-to-end 3D Neural Network for Very Sparse Multi-View Stereopsis	TPAMI 2021	paper code	★☆
16	Vis-MVSNet	Visibility-aware Multi-view Stereo Network	BMVC 2020 best paper	paper code	★★★★
17	PVSNet	PVSNet: Pixelwise Visibility-Aware Multi-View Stereo Network.	CVPR 2020	paper	★★★★☆
18	D2HC-RMVSNet	Dense Hybrid Recurrent Multi-view Stereo Net with Dynamic Consistency Checking	ECCV 2020	paper code	★★★
19	BP-MVSNet	BP-MVSNet: Belief-Propagation-Layers for Multi-View-Stereo	3DV 2020	paper	★☆
20	TransMVSNet	TransMVSNet: Global Context-aware Multi-view Stereo Network with Transformers	CVPR 2022	paper code	★★★☆
21	GBi-Net	Generalized Binary Search Network for Highly-Efficient Multi-View Stereo	CVPR 2022	paper code	★★★★★
22	IterMVS	IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo	CVPR 2022	paper code	★★☆
23	CIDER	Learning Inverse Depth Regression for Multi-View Stereo with Correlation Cost Volume	CVPR 2019	paper code	★★★
24	UniMVSNet	Rethinking Depth Estimation for Multi-View Stereo: A Unified Representation and Focal Loss	CVPR 2022	paper code	★★★★
25	AA-RMVSNet	AA-RMVSNet: Adaptive Aggregation Recurrent Multi-View Stereo Network	ICCV 2021	paper code	★★★
26	MVSTER	MVSTER: Epipolar Transformer for Efficient Multi-View Stereo	ECCV 2022	paper code	★★★☆
27	EPP-MVSNet	EPP-MVSNet: Epipolar-Assembling Based Depth Prediction for Multi-View Stereo	CVPR 2022	paper code	★★★
28	CDS-MVSNet	Curvature-guided dynamic scale networks for Multi-view Stereo	ICLR 2022	paper code	★★★
29	LANet	Long-range Attention Network for Multi-View Stereo	WACV 2021	paper	★★☆
30	ACMMP	Multi-Scale Geometric Consistency Guided and Planar Prior Assisted Multi-View Stereo	TPAMI 2022	paper code	★★★★★
31	CER-MVS	Multiview Stereo with Cascaded Epipolar RAFT	ECCV 2022	paper code	★★
32	NP-CVP-MVSNet	Non-parametric Depth Distribution Modelling based Depth Inference for Multi-view Stereo	CVPR 2022	paper code	★★★★☆
33	GigaMVS	GigaMVS: A Benchmark for Ultra-large-scale Gigapixel-level 3D Reconstruction	TPAMI 2021	paper code	★★★★
34	RayMVSNet	RayMVSNet: Learning Ray-based 1D Implicit Fields for Accurate Multi-View Stereo	CVPR 2022	paper	★★★☆
35	Effi-MVS	Efficient Multi-view Stereo by Iterative Dynamic Cost Volume	CVPR 2022	paper code	★★★

✒️基本概念

• 对极几何 | Epipolar Geometry
• 单应变换 | Homography
• 平面扫描 | Plan Sweeping
• 立体几何匹配 | Stereo Matching
• 深度图像 | RGB-D

🔬项目/代码

• fusibile深度图融合点云
• DTU定量测试
• MVSNet
• MVSNet_pl
• PointMVSNet
• CVP-MVSNet
• CasMVSNet

mvs's Issues

Tank&Temples深度图融合

同学你好，能请教下你对于Tank&Temples数据集，你在对每个场景做深度图融合时，都有去精调photo_threshold和thres_view参数吗？

参考图像

请问我如何能知道重建出的点云是以哪张图片作为参考图像的？谢谢您！

关于colMAP和MVS深度图估计与优化

学长您好，Colmap中代价构造、累积、估计和优化是封装在一起的，colMAP的重建方法是基于GEM模型的，那么我们使用MVS算法是优化了其深度图估计与优化步骤的内容吗？剩余内容一般是通过泊松表面重建和纹理映射优化的吗