Zi Jian Yew

Hi, the ModelNet40 dataset is supposed to be downloaded automatically from the PointNet server, but it seem that there is currently a server misconfiguration. I'll see if I can find...

The pretrained network most likely won't generalize to 2D point clouds. But from the algorithm point of view, I don't think there's anything stopping you from using 2D point clouds....

IMHO, the definitions are not strict, so both terms sometimes mean the same thing. However, registration typically refers to the problem of computing the parameters (e.g. rotation and translation in...

Hi, 1. You understood it wrongly, perhaps the description could have been written more clearly. We compute the loss for all (two) iterations, where each iteration's loss is the sum...

1. Yes, we select up to 64 points that within the desired radius as in the original PointNet++. This works better than using the closest points in practice. 2. The...

I think you are referring to the license? The source code uses a MIT license, as seen in the license file in the repository.

Oh, you mean the codes of the original RPM? The author's codes can be found [here](https://www.cise.ufl.edu/~anand/students/chui/tps-rpm.html). This version is written for non-rigid point cloud registration. For the results in my...

We perform the normalization in log space in the actual implementation for numerical stability. The operations are otherwise the same, e.g. the division in linear space is changed to subtraction...

Hi, 1. Sure, you can find my trained checkpoints of DCP [here](https://drive.google.com/drive/folders/18BDY_9phekmWmQcbhlFQQ-ZrLrg1AzHe?usp=sharing). I didn't check through the files, although these should be the correct files. But let me know if...

That's weird. `prepool` is just a simple MLP which shouldn't lead to nan's (unless the weights or inputs are nan's). I'm not sure what might be the cause since the...