Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

代码中的matlab问题 #54

Open
lgc1030 opened this issue May 6, 2023 · 10 comments
Open

代码中的matlab问题 #54

lgc1030 opened this issue May 6, 2023 · 10 comments

Comments

@lgc1030
Copy link

lgc1030 commented May 6, 2023

No description provided.

@SHERRYMINE
Copy link

  1. Change the path parameters in the code
  2. Run BaseEvalMain_web.m
  3. Run ComputeStat_web.m

@SHERRYMINE
Copy link

SHERRYMINE commented May 9, 2023

In BaseEvalMain_web.m:
1.dataPath: download Points(6.3G) from http://roboimagedata.compute.dtu.dk/?page_id=36
2.plyPath: Mvsnet output path
3.resultsPath: BaseEvalMain_web.m output path

In ComputeStat_web.m:
1.dataPath: keeps the same path in BaseEvalMain_web.m
2.resultsPath:keeps the same path in BaseEvalMain_web.m

@lgc1030
Copy link
Author

lgc1030 commented May 10, 2023

In BaseEvalMain_web.m: 1.dataPath: download Points(6.3G) from http://roboimagedata.compute.dtu.dk/?page_id=36 2.plyPath: Mvsnet output path 3.resultsPath: BaseEvalMain_web.m output path

In ComputeStat_web.m: 1.dataPath: keeps the same path in BaseEvalMain_web.m 2.resultsPath:keeps the same path in BaseEvalMain_web.m

您好,我跑ComputeStat_web.m的结果是mean acc 0.4866 mean comp 0.3943,请问这个结果是论文中提到的最后结果吗?我不太理解ComputeStat_web.m和BaseEvalMain_web.m跑出来的结果分别代表什么意思,那个结果是论文中提到的结果呢,麻烦您指教一下可以吗,谢谢

@SHERRYMINE
Copy link

你可以再跑一遍,注意看那两个matlab文件在命令行跑出来的结果。
mean acc是指平均准确率,mean comp是指平均完整率,具体的评价指标在论文中有写,表示使用mvsnet模型训练后融合生成的点云和ground truth点云的之间的差别。

直观上的理解是这样的,0.4866 mean acc可以理解成:在生成的点云中,平均有48.66%的点深度和gt是一样的(在可接受的范围内);0.3943 mean comp可以理解成:生成的点云数量平均占gt的39.43%

@SHERRYMINE
Copy link

去看Read me的Results on DTU

@SHERRYMINE
Copy link

Deep Learning for Multi-view Stereo via Plane Sweep A Survey
我想你看完这一篇论文就都会了

@lgc1030
Copy link
Author

lgc1030 commented May 16, 2023

Deep Learning for Multi-view Stereo via Plane Sweep A Survey 我想你看完这一篇论文就都会了

好的

@lgc1030
Copy link
Author

lgc1030 commented May 20, 2023

Deep Learning for Multi-view Stereo via Plane Sweep A Survey 我想你看完这一篇论文就都会了

您好,请问提供的matlab代码中,哪一个matlab文件跑出来的结果是论文当中的acc和comp呀

@SHERRYMINE
Copy link

SHERRYMINE commented May 20, 2023 via email

@ZooKangaroo
Copy link

@SHERRYMINE 请问如果我需要定量评价自己数据集的重建效果,需要怎么操作?如果使用这里给的matlab代码的话,那么obsmask文件是什么?该怎样生成?谢谢!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants