Missing ReadMe

[007-gif]

Hi,Can you write a Readme introduction file about how the code runs？I'm having a little run-time problem。

[AlexPop104]

Hello!

I made this repo to keep track of my work with a colleague. This more or less should have been private and only to be shown when clean (because "who would read dirty code which recreates a paper"). You are seeing in this repo all the work by days and tests.

Having said that, I can gladly help. Until now we split our repo in two parts, one where I was dealing with classification and the other which my colleague dealt with semantic segmentation. We are merging our code base now so we are all on the same page.

Bear in mind, we are two students with not much experience in coding trying to solve problems and iterate through new experiments.

We tried to recreate what Tegusi did with RGCNN and wrote very dirty code solutions just to iterate fast.

It will take some time until I write a reliable Readme because the focus was not so much on presentation as much as writing new code.

Until we start writing the Readme, I can gladly answer questions by email at alexandru.pop@aut.utcluj.ro

Best regards,

[007-gif]

My research is focused on Point cloud segmentation.I assume you're running TeGusi's github code without a problem.But I found at run time that on OneDrive, the only datasets were abel_val.npy and label_test.npy, but this is not enough when running code, data_val.npy, data_test.npy, label_train.npy, data_train.npy files are also required. For me, as a newcomer, it was difficult to convert ModelNet10 data into numpy data using PointNet. Do you still have the corresponding data?Or how can I generate the appropriate numpy files so that I can run the code successfully?

[DomsaVictor]

Hi. Yes, we were able to run the tensorflow implementation of RGCNN. Please email me at domsavictor@gmail.com and I will send you some code to create the .npy files.

[007-gif]

I have sent emails by jiaqiw007@gmail.com,please remember to check.Thank you very much for your help！