enyahermite / picassoplus Goto Github PK

Hi，
I Try many times to download the dataset through your link, but there is nothing in the cloud drive.
I would greatly appreciate it if you could share another effective link to obtain the dataset.
Thanks!

Dear Enya,
I'm a masters student from Technical University of Denmark who is trying to use your framework for a project of mine.
I don't have much experience with TensorFlow so I'm using your Pytorch implementation.
I'm running the PicassoNet-II on a GPU-cluster operating on on a Ubuntu 18.04.6 LTS GPU cluster and i encounter the following error:

terminate called after throwing an instance of 'thrust::system::system_error'
what(): CUDA free failed: cudaErrorInvalidAddressSpace: operation not supported on global/shared address space

I believe the error comes from the decimate_gpu.cu however I'm not entirely sure why or how.
Can you help with this one?

Best,
Bjonze

Hi @EnyaHermite, thanks for open-sourcing the code for this awesome work.
Could you please provide more information on how to process the datasets, such as scannet, scannet_texture, etc? I'm trying to run the pytorch branch of this repo.

Moreover, the gdrive link for the trained models is not working?

as title

Thank you for the great work! In point-based covolution, input point features are input in batches, rather than concat each sample like mesh-based covulution. This requires the number of vertices in each sample to be the same. But due to the triangle degeneration, the real vertex number for every cropped sample will slightly differ from each other. How did you solve the problem of the different number of vertices in each sample in point-based covolution. Is it to fill the vertices of each sample in the batch according to the desired vertex size in each resolution mesh?
Hungry for your reply, thank you very much!

Hi @EnyaHermite

I am attempting to execute your pytorch version of PicassoNetII as defined in shape_cls.py and it is FAR from being executable.

I have encountered numerous operations in which two inputs are on gpu and cpu, wrong argument count for DualBlock and EncoderMeshBlock, and class attributes that do not exist (named incorrectly).

I am unsure if I can use this code. It is clear the pytorch PicassoNetII as defined in shape_cls.py was never tested.

Hi, @EnyaHermite

When will the code be released?

Thanks~

nnquery_gpu.cu
D:\vs2019\VC\Tools\MSVC\14.29.30133\bin\HostX86\x64\link.exe /nologo /INCREMENTAL:NO /LTCG /DLL /MANIFEST:EMBED,ID=2 /MANIFESTUAC:NO /LIBPATH:D:\Anaconda\envs\picassoplus\lib\site-packages\torch\lib "/LIBPATH:C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.3\lib/x64" /LIBPATH:D:\Anaconda\envs\picassoplus\libs /LIBPATH:D:\Anaconda\envs\picassoplus /LIBPATH:D:\Anaconda\envs\picassoplus\PCbuild\amd64 /LIBPATH:D:\vs2019\VC\Tools\MSVC\14.29.30133\lib\x64 "/LIBPATH:C:\Program Files (x86)\Windows Kits\10\lib\10.0.19041.0\ucrt\x64" "/LIBPATH:C:\Program Files (x86)\Windows Kits\10\lib\10.0.19041.0\um\x64" c10.lib torch.lib torch_cpu.lib torch_python.lib cudart.lib c10_cuda.lib torch_cuda_cu.lib torch_cuda_cpp.lib /EXPORT:PyInit_point_nnquery build\temp.win-amd64-cpython-38\Release\picasso/point/pi_modules/source/nnquery.obj build\temp.win-amd64-cpython-38\Release\picasso/point/pi_modules/source/nnquery_gpu.obj /OUT:build\lib.win-amd64-cpython-38\point_nnquery.cp38-win_amd64.pyd /IMPLIB:build\temp.win-amd64-cpython-38\Release\picasso/point/pi_modules/source\point_nnquery.cp38-win_amd64.lib
正在创建库 build\temp.win-amd64-cpython-38\Release\picasso/point/pi_modules/source\point_nnquery.cp38-win_amd64.lib 和对象 build\temp.win-amd64-cpython-38\Release\picasso/point/pi_modules/source\point_nnquery.cp38-win_amd64.exp
nnquery.obj : error LNK2001: 无法解析的外部符号 "public: long * __cdecl at::TensorBase::data_ptr(void)const " (??$data_ptr@J@TensorBase@at@@QEBAPEAJXZ)
build\lib.win-amd64-cpython-38\point_nnquery.cp38-win_amd64.pyd : fatal error LNK1120: 1 个无法解析的外部命令
error: command 'D:\vs2019\VC\Tools\MSVC\14.29.30133\bin\HostX86\x64\link.exe' failed with exit code 1120

I don't konw how can I resolve this question,I think .cu files need modify by some ways