Pointnet2.PyTorch icon indicating copy to clipboard operation
Pointnet2.PyTorch copied to clipboard

A faster implementation of PointNet++ based on PyTorch.

Results 34 Pointnet2.PyTorch issues
Sort by recently updated
recently updated
newest added

When I build .sh file. This string occur. but no build error. I use gcc 4.9, cuda-9.0 Thanks ahead.

Hi there, recently I upgraded the pytorch to 1.4.0 and the forward function of pointnet2 broke with errors like: `CUDA: an illegal memory access was encountered.` or `CUDA kernel failed...

in the file: tool/pointnet2_msg, line 69, pc[..., 3:].transpose(1, 2).contiguous(). why need transpose(1, 2)?

@sshaoshuai 您好,非常感谢您提供的代码,我使用的时候出了一些问题,想向您请教一下。我成功地执行了 python setup.py install 代码,CUDA的OP也可以成功调用,但是训练过了几个epochs就会出现一个问题:CUDA: an illegal memory access was encountered. (每次训练的报错行数不一样,但是最后的报错一样)经过排查我觉得可能与CUDA和里面的线程分配有关,而且这个代码我之前几个月用着都没问题,这个问题是最近出现的。请问您是否遇到过类似的情况?下面的是一次报错的具体信息。非常感谢。 `e=/pytorch/aten/src/THC/THCCachingHostAllocator.cpp line=265 error=77 : an illegal memory access was encountered THCudaCheck FAIL file=/pytorch/aten/src/THC/THCCachingHostAllocator.cpp line=265 error=77 : an...

Hi @sshaoshuai, Shouldn't evaluation be performed in `eval()` mode? https://github.com/sshaoshuai/Pointnet2.PyTorch/blob/master/tools/train_and_eval.py#L103

you give the task of KITTI ourdoor foreground point cloud segmentation example, so what's the result can be referenced?

Hi Shaoshaui, Regarding the network architecture in your pytorch implementation. I noticed that in the SA and FP modules, the mlp / conv2d channel input and output dimensions differ from...

Could you please give an example of how to use your code within Python? I have several lists of 3D points, which each represent an object.

Hi, I have a few questions about this repo. 1. How much faster is your implementation compare to the original pytorch/TensorFlow version ? 2. Can you explain more about how...

![image](https://github.com/sshaoshuai/Pointnet2.PyTorch/assets/73287606/57e45e7f-40c4-4dee-a5ab-76a183370191) Hello, what I should dwonload? every link?