Open MarquessMar opened 7 years ago
z1042511771
commented
5 years ago Sorry to bother, but I'm confused about the format of the data loaded in load_data.py, could you please tell me about how many dimensions the loaded data should have and what each dimension represents? Thanks a lot.
Did you solve the problem? I have the same issue with you
MarquessMar
commented
5 years ago not quite.
I modified the code and loaded the data in my own way. Below is how I do it. It's better if I could just send you my code, but unfortunately I have already deleted it because I gave up on using this project.
in main.py, you can see that the loaded data should be 4D, organized in the same order as the default tensorflow format, which is [batch, height, width, channel].
therefore, i write a MATLAB code to save a .mat file which includes four variables, horizontal, vertical, 4views, and singleview. For one image, concatenate different views in the channel axis, and concatenate all the images in the batch image, then you will get a 4D variable which could be a tensorflow tensor directly. After that, you only need to read the .mat file using scipy.io.loadmat function, and pass them directly into the network.
Hope this helps. If you have other questions, feel free to email me.
Good luck on your research!
Kind regards, -- Haoxin Ma Mechanical Engineering, Tsien Hsue-Shen Honored Program School of Mechanical Engineering Shanghai Jiao Tong University Tel: +86 132 6299 1801
----- 原始邮件 ----- 发件人: "z1042511771" notifications@github.com 收件人: "youngjinYoon/SPL_LF_SR" SPL_LF_SR@noreply.github.com 抄送: "MarquessMar" marquess@sjtu.edu.cn, "Author" author@noreply.github.com 发送时间: 星期四, 2019年 3 月 28日 上午 11:38:16 主题: Re: [youngjinYoon/SPL_LF_SR] A little problem using load_data.py (#3)
Sorry to bother, but I'm confused about the format of the data loaded in load_data.py, could you please tell me about how many dimensions the loaded data should have and what each dimension represents? Thanks a lot.
Did you solve the problem? I have the same issue with you
-- You are receiving this because you authored the thread. Reply to this email directly or view it on GitHub: https://github.com/youngjinYoon/SPL_LF_SR/issues/3#issuecomment-477434923
Sorry to bother, but I'm confused about the format of the data loaded in load_data.py, could you please tell me about how many dimensions the loaded data should have and what each dimension represents? Thanks a lot.