lvjiahui / drt Goto Github PK

Hi Jiahui,

I am trying to run one of the datasets, specifically the horse. However, my local graphics card is really out-date and is not able to handle the whole dataset including 72 images. Is it possible just using half of them? I noticed that your code specifies the number of images as 72. Is there any way modifying it to adjust to 36 images? Thank you!

Here is part of the error message:
RuntimeError: CUDA out of memory. Tried to allocate 16.00 MiB (GPU 0; 3.95 GiB total capacity; 2.80 GiB already allocated; 14.94 MiB free; 2.80 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

Hi, Jiahui. Is the transparent objects with the single refractive index manufactured by your lab? Because the transparent objects with single refractive index in market are rarely found (Mainly surface is covered with a layer of glaze).

Hi, jiahui,

Is the environment matting code involved in this repo? I found a matlab env matting code in the siggraph 2018 paper, is it the same as in your paper?

I've download NVIDIA-OptiX 6.5.0, and edited config.py for the path to library. However, when running optim.py, an error occurred like:

And here is my file lists:

I wonder if I need to install Optix? (I've searched for it but can't get a uniform solution)

Hi, jiahui, when I run optix.py the error is “ImportError: No module named 'optix'” ,and I set the absolute Optix path in config.py.

the detail error information:

Traceback (most recent call last):
File "/home/qining/CODE/DRT-master/optim.py", line 6, in
import DiffRender as Render
File "/home/qining/CODE/DRT-master/DiffRender.py", line 6, in
extra_include_paths=[optix_include], extra_ldflags=["-L"+optix_ld, "-loptix_prime"])
File "/home/qining/anaconda3/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 980, in load
keep_intermediates=keep_intermediates)
File "/home/qining/anaconda3/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 1196, in _jit_compile
return _import_module_from_library(name, build_directory, is_python_module)
File "/home/qining/anaconda3/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 1543, in _import_module_from_library
file, path, description = imp.find_module(module_name, [path])
File "/home/qining/anaconda3/lib/python3.6/imp.py", line 298, in find_module
raise ImportError(_ERR_MSG.format(name), name=name)
ImportError: No module named 'optix'

and my config.py:

optix_include = "/home/qining/NVIDIA-OptiX-SDK-6.5.0-linux64/include"
optix_ld = "/home/qining/NVIDIA-OptiX-SDK-6.5.0-linux64/lib64"

Would you give me some suggestions?Thank you!

Hi Jiahui,

Sorry I am just a beginner in the field of computer graphics and lacking relevant background knowledge.

In the paper, you mention that you use method from Kutulakos and Seitz (2020) to reconstruct the initial rough shape. I don't read this paper thoroughly but it seems their method produces a voxel-based model? How this result could be used in your mesh-based method?
Also, it would be nice if you could offer relevant codes about reconstructing the shape using silhouette images.

Thank you!

My Environment: ubuntu 18.04; pytorch 1.3.1

Traceback (most recent call last):
File "optim.py", line 6, in
import DiffRender as Render
File "/home/Code/DRT/DiffRender.py", line 7, in
extra_include_paths=[optix_include], extra_ldflags=["-L "+optix_ld, "-loptix_prime"])
File "/home/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 661, in load
is_python_module)
File "/home/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 841, in _jit_compile
return _import_module_from_library(name, build_directory, is_python_module)
File "/home/anaconda3/envs/pytorch/lib/python3.6/site-packages/torch/utils/cpp_extension.py", line 1052, in _import_module_from_library
return imp.load_module(module_name, file, path, description)
File "/home/anaconda3/envs/pytorch/lib/python3.6/imp.py", line 243, in load_module
return load_dynamic(name, filename, file)
File "/home/anaconda3/envs/pytorch/lib/python3.6/imp.py", line 343, in load_dynamic
return _load(spec)
ImportError: liboptix_prime.so.6.5.0: cannot open shared object file: No such file or directory

Hi jiahui,

Thank you for making the code public, it helps me a lot . I have a questions about the dataset, I could not find the capture pictures in the dataset, is it include in your public data set? because I want to build the visual hull use your data set, would you mind to upload it ?Thank you!

when i run "python optim.py"，i encountered the following error:
Traceback (most recent call last):
File "optim.py", line 224, in
scene = optimize(HyperParams)
File "optim.py", line 175, in optimize
scene = Render.Scene(f"{config.data_path}{name}_vh.ply")
File "/DRT-master/DiffRender.py", line 301, in init
self.update_mesh(mesh_path)
File "/DRT-master/DiffRender.py", line 313, in update_mesh
self.optix_mesh.update_mesh(opt_F, opt_v)
RuntimeError: Function "RTPresult _rtpModelUpdate(RTPmodel, unsigned int)" caught exception: Encountered a CUDA error: cudaEventRecord( m_eventEnd, stream ) returned (700): an illegal memory access was encountered

The environment I configured should be correct. How can I solve this problem? Is it related to the cuda version