During running the script my_train.py, I get this error: -

Traceback (most recent call last):
  File "./scripts/my_train.py", line 14, in <module>
    from lib.module.PNSPlusNetwork import PNSNet as Network
  File "/content/drive/Shareddrives/VPS/VPS_main/lib/module/PNSPlusNetwork.py", line 11, in <module>
    from module.PNSPlusModule import NS_Block
  File "/content/drive/Shareddrives/VPS/VPS_main/lib/module/PNSPlusModule.py", line 10, in <module>
    import self_cuda_backend as _ext
ModuleNotFoundError: No module named 'self_cuda_backend'

Now,, I did run build as instructed in the readme and it generates a lot of log, but this error appears over and over again: -

/content/drive/Shareddrives/VPS/VPS_main/lib/module/PNS
running build
running build_ext
/usr/local/lib/python3.7/dist-packages/torch/utils/cpp_extension.py:387: UserWarning: Attempted to use ninja as the BuildExtension backend but we could not find ninja.. Falling back to using the slow distutils backend.
  warnings.warn(msg.format('we could not find ninja.'))
/usr/local/lib/python3.7/dist-packages/torch/utils/cpp_extension.py:788: UserWarning: The detected CUDA version (11.1) has a minor version mismatch with the version that was used to compile PyTorch (11.3). Most likely this shouldn't be a problem.
  warnings.warn(CUDA_MISMATCH_WARN.format(cuda_str_version, torch.version.cuda))
building 'self_cuda_backend' extension
x86_64-linux-gnu-gcc -pthread -Wno-unused-result -Wsign-compare -DNDEBUG -g -fwrapv -O2 -Wall -g -fstack-protector-strong -Wformat -Werror=format-security -g -fwrapv -O2 -g -fstack-protector-strong -Wformat -Werror=format-security -Wdate-time -D_FORTIFY_SOURCE=2 -fPIC -I/usr/local/lib/python3.7/dist-packages/torch/include -I/usr/local/lib/python3.7/dist-packages/torch/include/torch/csrc/api/include -I/usr/local/lib/python3.7/dist-packages/torch/include/TH -I/usr/local/lib/python3.7/dist-packages/torch/include/THC -I/usr/local/cuda/include -I/usr/include/python3.7m -c PNS_Module/sa_ext.cpp -o build/temp.linux-x86_64-3.7/PNS_Module/sa_ext.o -std=c++11 -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE="_gcc" -DPYBIND11_STDLIB="_libstdcpp" -DPYBIND11_BUILD_ABI="_cxxabi1011" -DTORCH_EXTENSION_NAME=self_cuda_backend -D_GLIBCXX_USE_CXX11_ABI=0
In file included from /usr/local/lib/python3.7/dist-packages/torch/include/torch/extension.h:4:0,
                 from PNS_Module/sa_ext.cpp:2:
/usr/local/lib/python3.7/dist-packages/torch/include/torch/csrc/api/include/torch/all.h:4:2: error: #error C++14 or later compatible compiler is required to use PyTorch.
 **#error C++14 or later compatible compiler is required to use PyTorch.**

I feel like build needs to run properly for the whole code to execute well. The problem is on Colab, the CUDA version is not 10.0 and there is no way to install C++ there. On my local machine, I only have support for Colab 11.0 and above. Is there any way to restructure this code for CUDA 11.0 or just run it using CPU?

Hi,
I am trying to download your annotations but the OneDrive link has expired and I cannot use the other link.
Could you please update it ?
Thank you

编译时遇到这个错误该如何解决？

hi,
First of all, thank you for your excellent work.
when I try to train, I get the following error:

ImportError: self_cuda_backend.cpython-36m-x86_64-linux-gnu.so: undefined symbol _ZN6caffe26detail36_typeMetaDataInstance_preallocated_7E.

And this is reported when running the PNSPlusMoudle.py.
My OS is centos7, Cuda version is 10.0, PyTorch version is 1.1, GPU is titan XP with 12GB memory.

执行python ./scripts/my_train.py会报以下错误：
CUDA error at: PNS_Module/sa_ext.cpp:34
no kernel image is available for execution on the device cudaGetLastError()

Thanks for reading this question!
The email address provided at the end of the data download page of the SUN Colonoscopy Video Database is hitoh (a t) mori.m.is.nagoya-u.ac.jp,
This is an unfeasible email address, may I directly change the email address to [email protected] and then send the request email?
thank you!

Thank you very much for making this dataset available. We sent our application email to [email protected], but it was returned. Therefore, we would like to confirm if the email address for submitting applications is correct.

Please ask the “sh . /utils/reorganize.sh” in step 3 is executed in which directory？

hi I have followed [Upgrade environment for NS block.pdf] from step1 to step5 but it still doesn't work. How can I solves this problem?
thanks

Hi, your work is respectable！But I have some problems. Could you please help me?
when i type
pip install git+https://github.com/pytorch/tnt.git@master
there are some errors occured.

By the way, when I run python setup.py build develop, I also have some problmes.

for idx, (img, label) in enumerate(zip(img_li, label_li)):
if idx == 0:
IMG = torch.zeros(len(img_li), *(img.shape))
LABEL = torch.zeros(len(img_li) - 1, *(label.shape))
IMG[idx, :, :, :] = img
else:
IMG[idx, :, :, :] = img
LABEL[idx - 1, :, :, :] = label

so, why LABEL has less frame than IMG ? i 've been confused about it ?

Hello, the data set format is formatted according to the requirements, but this error occurs

File "./lib/dataloader/dataloader.py", line 32, in
int(name.split('-')[0].split('_')[-1]),
ValueError: invalid literal for int() with base 10: '00003.jpg'

I'm not sure what that is

Hello author! I can't find the metrics results or prediction maps of the image-based models on SUN-SEG-Easy/Hard Seen datasets. Can you provide the prediction maps or the trained weights? Thanks!

Hello, I ran 15 epochs with the code you gave, and the generated pth file was used to run the test set, but the test results were not as good as the pth path you gave, why is this?

Hello, I am a newcomer to video segmentation and want to try the video segmentation project, is there any baseline that is more suitable for novices?

为什么运行python setup.py build develop会出现语法报错？？？我并没有更改它，而是直接从github下载得到的。

Sorry to bother you, but regarding the one-key evaluation toolbox mentioned in the paper, the address provided 'https://github.com/GewelsJI/VPS/tree/ main/eval' is invalid. Where can I obtain it?

Hello, I observed that the code provides some evaluation indicators, in the path ' ./eval/eval-result', where maxDice and maxIou represent the maximum value or the average value? Notice that there is this code in the metrics.py: column_Iou = np.mean(self.threshold_Iou, axis=0).Looking forward to your reply!

Hi, if i download dataset from the link drive,there are some files are missing after the download..

First of all, thank you for this meaningful job！
Secondly, for the Visual Attributes you proposed, if I want to get the Visual Attributes corresponding to each picture, could you tell me where I should look for it?
e.g. image_name_00001.jpg --> IB , HO
image_name_00002.jpg --> FM , SO
......

Hi, sorry to bother you. I have some questions about your dataset, and I have tried to contact the author of the SUN dataset but got no response. So I still take the liberty to ask you.

It is mentioned in your document(https://github.com/GewelsJI/VPS/blob/main/docs/DATA_DESCRIPTION.md):
As such, it yields the final version of our SUN-SEG dataset, which includes 49,136 polyp frames (i.e., positive part) and 109,554 non-polyp frames (i.e., negative part) taken from different 285 and 728 colonoscopy videos clips, as well as the corresponding annotations. The following sections will provide details about our SUN-SEG point-by-point.

Are there only frames containing polyps in your videos in the positive part? And how do you use the non-polyp frames in your train
and test dataset? How do you deal with the non-polyp frames when you calculate evaluation such as Dice and so on?

Thank you very much and look forward to your reply.

I downloaded the benchmark file, but I don't know how I should use it. Could you please tell me? Thanks!
“We provide an out-of-the-box evaluation toolbox for the VPS task, which is written in Python style. You can just run it to generate the evaluation results on your custom approach. Or you can directly download the complete VPS benchmark including prediction map of each competitor at download link: OneDrive / Baidu Drive (Password: 2t1l, Size: 5.45G).”

Hi, thanks to your nice work.

I am leaving a message because I have a question.
First, I want to find the fps. Second, I have an actual endoscopy video, and I want to make it into a video that shows the location of the lesion in real time, like your video. Is there any related code?

Hi, QR code not available, please update.

你好，使用linux编译会报错怎么解决？

Hello, I failed to get the SUN data set and failed to send the email. Could you please share the data set or help correct the email address? Thank you

Hello, I wonder where to place the test model(.pth) for the ./eval/vps_evaluator.py? I tried putting it under ./eval and ./VPS-MAIN, named 2022-TMI-PNSPlus.pth, but found that the evaluation result has nothing to do with my model. Looking forward to your reply, thanks a lot.

二维码已过期，请问有何方法进入到您的微信群

Hello, now I have something I don't understand about the NS module in the model, and I would like to ask you. If the role of the first NS module is to fuse local and global features, then what is the role of the second NS module, I look forward to receiving your reply.

Thanks for your great work. I have contacted you by email ([email protected]). Can you provide me the link to download the dataset as soon as possible?

"python setup.py build develop"instructions can run, "self_cuda_backend"cannot be imported in PNSPlusModule.py, PNSPlusNetwork.py also cannot be Compiled