error report in python version

[rsy6318]

Hi, I found that in some pairs of source and target point clouds, there exists an error.
Looking forward to your reply!
(pugeo) siyuren_21@ss420f:~/data1/siyuren_21/3D-Registration-with-Maximal-Cliques-main$ /home/siyuren_21/anaconda3/envs/pugeo/bin/python /home/siyuren_21/data1/siyuren_21/3D-Registration-with-Maximal-Cliques-main/main_rmc_1.py Graph construction: 12.66ms Search maximal cliques: 1973.51ms Total: 41546 After filtered: 4838 Traceback (most recent call last): File "/home/siyuren_21/data1/siyuren_21/3D-Registration-with-Maximal-Cliques-main/main_rmc_1.py", line 344, in <module> pred_trans,pred_trans1,spend_time=test(src_path,tgt_path) File "/home/siyuren_21/data1/siyuren_21/3D-Registration-with-Maximal-Cliques-main/main_rmc_1.py", line 283, in test batch_A = src_pts[list(macs[group[i][0]])][None] IndexError: list index out of range

[rsy6318]

Here is the error reported.

[rsy6318]

Here is the point clouds I have used, which is produced by voxel downsample from 3DMatch dataset.
src_tgt_trans.zip

[zhangxy0517]

Please also provide the input correspondence data.

[rsy6318]

There is no input correspondence data. I use Open3d to calculate FPFH features and then estimate initial correspondence. Here is my test code, which is modified from example_fpfh_fcgf.py.
main_rmc_1.zip

[zhangxy0517]

If you plan to use downsampled sparse point cloud (5000 points) to calculate FPFH features, downsample should be set to a much bigger value (e.g. 0.25) to ensure FPFH can compute features for the sparse points correctly.

[zhangxy0517]

The bug is fixed, please download the newest code.

[rsy6318]

Thanks ! I would like to try it !

[rsy6318]

I have run the fixed python code on 3DMatch dataset and get the results "RE: 2.26 TE: 7.14cm", which is different from the results on the paper. Is it normal? Or could you provide your registration results?

[rsy6318]

Here is my impliment details. First, we downsample the original point clouds wich voxel size 0.05 and store them. When testing, we calculate the FPFH features and run your fixed codes to get the registration results.