gaolinorange / automatic-unpaired-shape-deformation-transfer Goto Github PK

I got a following issue while running main.py file.

vagan.tb = FLAGS.tb

AttributeError: 'Namespace' object has no attribute 'tb'

Thank you

Hi

First of all, thank you for providing the source cod publicly available,
in the ReadMe file in folder ACAP_linux you said we must to modify the variable ${GUROBI_ROOT} in CMakeLists.txt

What is the exact mean of your statement?
How I must modify it?

Thanks in advance

Hello, Dr. Gao, congratulations for your nice work.

I'm trying to run your code with my own obj files but I don't know where to start.

Could you tell me the steps how to run the code?

Thank you!

Thank you very much for the code provided by the author, but I have a small problem with the MATLAB script part, that is where does the file with the suffix ".art, .fdm, .cir, .ecc" required in LoadFeature. M come from?

Thank you for this work. Can you provide an example .mat (cat and horse etc), no need for the full dataset. I want to debug the code without the need to run the feature generation codes in matlab/windows. Thanks

1. LoadFeature.m
   I use metric(modelfolder_a,modelfolder_b) in matlabscript with my own model A and B to get the lightfield.mat as ABlfd.mat(if this was right), and got error:

Subscripted assignment dimension mismatch.
in LoadFeature (line 39)
FeatureVector(i,posArt) = LoadArt(artfile);

as well as the following 3 lines LoadFD, LoadCir and LoadEcc
although results of ndims() on both sides of the ‘=’ are equal, so I couldn't figure out where went wrong.

2. model.py
   I commented out 4 lines mentioned above and got ABlfd.mat， with togFeaturepoolingc.mat from get_feature17() for both A and B， run “python main.py --dataname_a A -dataname_b B” and got error:

..\python\model.py", line 125, in init
name='a/L1')
ValueError: Cannot create a tensor proto whose content is larger than 2GB.

Is this error possibly connected to the 4 lines I commented out? Those 4 lines seems only caculate the FeatureVector without changing size.

I'm new in matlab and tensorflow and may misunderstood these code. Any suggestion appreciated.

Dear Authors,
Congratulations on the beautiful work.
Is there a Linux version of the ACAP code available?
Regards,

Hi,

I am currently trying to train VC-GAN using my custom objects files.
However I encountered several problems.

1. Is the *lfd.mat file necessary to train the VC-GAN? it seems like it works without having the file present.
2. If *lfd.mat is needed, is there a python or matlab code to get the *lfd.mat?
3. When I tried training the model, it gave an error when computing the loss. It seems like the dimension does not match when adding two results from the self.metric_net. It show the following error: InvalidArgumentError (see above for traceback): Incompatible shapes: [46,1] vs. [17,1].
   The feature a and b sizes are (46,6294,3) and (17,3026,3), respectively.
   The random a and b sizes are (46,128) and (17,128), respectively.
   And the Ilf size is (128,1)
4. Is there any constraint for the obj files?
5. How much data do we need to train the network sufficiently?

I am running the code using python=2.7 and tf1.10

Thank you

Dear authors,

Thanks for providing an open source implementation for your interesting paper.

Please would it be possible to provide the pre-trained weights as well?

Many thanks in advance!

Best regards,

CoMISo-master依赖库巨多啊，配的我吐血了。

windows那部分代码是生成source与target数据集吗？

Hi, I was going through the paper and wondered what kind of dataset did you used for Fig. 25 and Fig.26.
Is it something that we can download online like MPI DYNA?

CoMISo的依赖多到爆炸啊，cmake不出来。
想问下

这些应该都是在windows上配置的吧，windows上的代码最后生成是什么？source与target的mesh文件吗？
linux上要跑的只是VC-GAN网络是吗

First of all, I have to say that I'm a fan of your method. I recently crossed the following article in the same spirit as yours

Yusuke Yoshiyasu, Wan-Chun Ma, Eiichi Yoshida, and Fumio Kanehiro. "As-Conformal-As-Possible Surface Registration." Computer Graphics Forum. Vol. 33. No. 5. 2014

I'm asking this question because the article was not cited in your work. How do you compare your ACAP to theirs?

Thank you again