zcong17huang/fastlfnet

fastlfnet's Issues

关于codeStep2训练的一点小问题

作者大大，在codeStep2训练时发现HCIloader中有一个
edge_name = os.path.join(LF_dir, 'BinaryEdge.png')
我在官方数据集和EPI数据集的这个路径中都没有找到这个图片文件，请问您这是在哪一步生成的嘛或者我哪一步漏了什么关键步骤嘛，谢谢解答！

training code and checkpoint

训练代码和训练好的权重可以提供吗

where is 'additional_invalid_area' ?

Hi, Thanks for your great work.

However, when I run codestep1 python train.py , it went wrong with this line

FastLFnet/codeStep1/train.py

Line 81 in 31d36ad

boolmask_img6 = imageio.imread(os.path.normcase(os.path.join(args.datapath, 'full_data/additional_invalid_area/kitchen/input_Cam040_invalid_ver2.png')))

It seems that I didn't have this image，Do I download the wrong dataset of kitchen scene? I only have additional folder data, but not additional_invalid_area and input_cam040_invalid_ver.png

训练的一些小问题

请问在训练时这个3出现了错误index 3 is out of bounds for axis 2 是什么原因呢，有点没看懂这个语句的意思，恳请大佬解答，感谢！

codeStep1 训练无法启动的问题

您好，不好意思又来打扰了，我按照您readme的步骤进行训练，输入命令：
python train.py --epochs 1000 --datapath D:\pycharm\light-field-depth\FastLFnet-main\dataset --savemodel D:\pycharm\light-field-depth\FastLFnet-main\demo --nums 9 --maxdisp 16

程序没有报错但是一直卡在第一步没有往下进行，如图示：

我只改动了dataloader的num_workers=4,请问您是否遇到过这样的问题？感谢您的解答！

关于训练卡住debug后的一些问题

昨天调试了一通程序，发现程序卡在下图中语句没有往下进行，初步猜测可能是数据读取有问题，遍历程序发现某些地方的数据设置和现有数据集不同，请问下作者大大是否数据集后期做了更新，可否更新一下您的数据设置方式或者在github中更新dataset呀，感谢感谢！（打扰多次非常不好意思）

About the One scale model problem?

HI, what is the one Scale model in Table 2? How do you construct the One scale model?
Can you share the One scale model code ?
Thks.

about codeStep1 “full_data/additional_invalid_area”

Thank you very much for your contribution to the field of deep learning. As a beginner, I ran the code after downloading hCI data set and found that the folder "full_data/additional_invalid_area" in codeStep1 did not appear in the data set I downloaded. Is this pre-processed or is there another step or link to get "additional_invalid_area"?