Some minor issues

[usherbraty]

1. In img_read_save.py, imsave(os.path.join(savepath, "{}.png".format(imagename)), image) should replace with imsave(os.path.join(savepath, "{}.png".format(imagename)), image.astype(np.uint8));
2. In test_IVF.py and test_MIF.py, you should add from torch.nn.parallel import DataParallel;
3. In README.md, the title of second output is wrong;
4. In the chart of paper, parameter Qbaf should be Qabf.

[Zhaozixiang1228]

Thank you very much for your attention to the project and thorough reading. Your suggestions are absolutely correct, and I will make improvements in the revised version.

[123456gwp]

Is there an incorrect subscript in Formula 5 of Section 3.3 in your paper? Whether V was mistakenly written as I.

[Zhaozixiang1228]

@123456gwp Yes, you are right, I wrote it incorrectly.

[windlu-1]

2. from torch.nn.parallel import DataParallel

imsave(os.path.join(savepath, "{}.png".format(imagename)), image.astype(np.uint8)); 由于使用了浮点数，所以定义SSIM时，需要加上data_range? data_range = image_F.max() - image_F.min() return ssim(image_F,image_A,data_range=data_range)+ssim(image_F,image_B,data_range=data_range)

[Gunkill]

请问M3FD数据集是如何划分训练集与验证集的呢