Out of memory when testing 3DMatch Dataset with FPFH descriptors.

[bobby1125]

Sorry to bother you! But when I test it on the 3DMacth Dataset with FPFH descriptors, the code is always out of memory in a computer with 48GB RAM at the following loop:
macs = graph.maximal_cliques(min=3) ... clique_weight = np.zeros(len(macs), dtype=float) for ind in range(len(macs)): mac = list(macs[ind]) if len(mac) >= 3: for i in range(len(mac)): for j in range(i + 1, len(mac)): clique_weight[ind] = clique_weight[ind] + SCG[mac[i], mac[j]]
What is the problem? Any hyperparameters need to be adjusted? By the way, we also find that the FCGF descriptors lead to a quite long time cost? Is it normal? FYI, all the correspondences are generated based in the code from SC2-PCR or PointDSC, and following the parameter setups in both your an their papers.

[zhangxy0517]

The Python version we provide is for demonstration purposes only. If you intend to conduct experiments with it, you will need to refer to the C++ code for additional implementation. The issue you mentioned likely occurs when there is a high inlier ratio in the input matches. In such cases, it is necessary to reduce the compatibility graph (line 1065-1228 in Linux/registration.cpp)

[bobby1125]

The Python version we provide is for demonstration purposes only. If you intend to conduct experiments with it, you will need to refer to the C++ code for additional implementation. The issue you mentioned likely occurs when there is a high inlier ratio in the input matches. In such cases, it is necessary to reduce the compatibility graph (line 1065-1228 in Linux/registration.cpp)

Many thanks to your reply! We will try to conduct experiments on the c++ version. By the way, we are wondering how you record or compute the memory consumption in the Table 13 of your paper?

[bobby1125]

Sorry to bother you again! But we find that the C++ version also runs out of memory when testing our generated 3DMatch(FPFH/FCGF/Geotransfoemer) and KITTI (FPFH) datasets with a 16GB RAM laptop. Is it a common problem?

[jeff0908]

Sorry to bother you again! But we find that the C++ version also runs out of memory when testing our generated 3DMatch(FPFH/FCGF/Geotransfoemer) and KITTI (FPFH) datasets with a 16GB RAM laptop. Is it a common problem?

@bobby1125
Hello here i have some experience about this question.
In my personal experience Geotransformer needs at least 13GB GPU ram to support its program running on our computer.

hope my experience can help u!!