tangeroo

compared with CA version2:deepFill v2

1

Hi, Sorry to bother.But I'm really curious about which is better when compared with deepFill v2.Have you ever tested that?

the result of my data seems very bad

3

Hi, I took sfm and stereo matching steps to obtain the camera poses and the depth maps. The poses were calculated by openmvg and the depth maps were calculated by...

suitesparse's version

4

hi, Sorry to bother again, but I wonder which version suitesparse you use in this project. I configure suitesparse5.1.2 in my system and get a libcholmod.so.3.0.11. Here the cmd points...

Can I use this project with windows g++ compiler?

2

@yanqingan Hi, Thanks for your excellent work. I don't have a linux system. I wonder if I can use this project with a wondows g++ compiler, like in MinGW. I'm...

parameter setting and other experience

4

Hi, Thanks for your sharing,it really helps.The ideas in this net are really great. I wonder which parameters I may set again if I want to use this net in...

openmvs和colmap里面的密集匹配算法异同

1

您好， 打扰了，又是从您的博客过来。不知您是否研究过这两个库里面的密集匹配算法，想请教一点问题。。最近我阅读了colmap密集匹配算法的文章，数学公式太多有些地方没怎么看明白，只知道colmap里面的方法view selection和depth estimation是通过EM迭代同时完成的，view selection考虑更为全面一些，等等。还想请教一下下面的问题： （1）colmap里面的深度值是怎么传播的，如何理解里面的状态转移方程 ，能否对比openmvs里面的方法稍微讲解一下不同之处。 （2）colmap法线优化是生成了额外的超参然后计算谁是最优，相比openmvs里面使用的二分法，这种方法的优势和劣势是什么 （3）相比openmvs，colmap方法更为耗时，耗时主要是因为EM这个迭代过程需要反复进行吗？ 国庆期间还发来问题，希望不会过分打扰到您。 谢谢~

Hi, Thanks for your work. I tested this program on an outdoor building data and get this problem. Can any tips about it? Thank you. 

Hi, Sorry to bother. I noticed that when big images input to the network, you set the overlap margin=8*up_scale. How did you set this margin. Recently I tested a SR...