pranavkdm/r-pointhop

R-PointHop: A Green, Accurate and Unsupervised Point Cloud Registration Method

r-pointhop's Issues

How to train own data?

Could you tell me how to train the network using my own data ?
Just change the train_data in the train.py ?

how much memory used while trainning

my pc with 16GB RAM can not run the train. It can not return from 'knn' function in point_utils.py
I have changed it into not using threading and changed the batch_size to smaller. It still not works.

What kind of equipment do you use? I always get stuck in point_utils.py 142 t.join()

How to reproduce the results on the partial point clouds?

Hi,

Thank you for releasing your codes, and it is a really excellent job! I wonder to reproduce the results on the partial-to-partial registration data, but I fail to find the settings of this experiment, such as the numbers of sample points and the numbers of the neighbors. Besides, I have no idea about that if the model needs to be re-trained on the partial data, or the released model can be used to register the partial inputs directly. Hope you can answer my doubts.

When I run the project encountered some problems

python train.py --first_20 False
Traceback (most recent call last):
File "train.py", line 51, in
main()
File "train.py", line 42, in main
train_data = train_data[train_label<20]
IndexError: boolean index did not match indexed array along dimension 1; dimension is 1024 but corresponding boolean dimension is 1

python test.py --source ./data/source_0.ply --target ./data/target_0.ply
munmap_chunk(): invalid pointer
(core dumped)

Is it because of the version of the installation library?

Could you tell me about your library version?

cannot get good result with custom data

I tried R-PointHop with custom data, but cannot get a good result.

How to reproduce the test results in the paper

I trained the network following the README Training part.
I am wondering how to reproduce the test results in the paper.

How do I resolve the following error when testing my own data set: IndexError: index 1024 is out of bounds for axis 2 with size 1024

I was able to test the test data provided in the source code, but when I replaced the test data with my own, there was an IndexError. What can I do about this?

When to update data with 3dmach dataset code

Hi,

Thank you for releasing your codes, and it is a really excellent job!I noticed that you and your partner published an updated version of this paper in March 2022, which contains 3dmach dataset, but the code of the paper has not been updated. Can you publish the new code？Thank you again.