jianzhangcs / ista-net-pytorch Goto Github PK

Using other dataset and mask retrain results poorly what changes need to be made in terms of code

Hi, when I run the test ISTA-Net+ for MRI_CS I get the flowing result:
MRI CS Reconstruction Start
CS ratio is 50, Avg Initial PSNR/SSIM for Brainimages_test is nan/nan
CS ratio is 50, Avg Proposed PSNR/SSIM for Brainimages_test is nan/nan, Epoch number of model is 200
MRI CS Reconstruction End
/home/amax/anaconda3/envs/pytorch/lib/python3.8/site-packages/numpy/core/fromnumeric.py:3372: RuntimeWarning: Mean of empty slice.
return _methods._mean(a, axis=axis, dtype=dtype,
/home/amax/anaconda3/envs/pytorch/lib/python3.8/site-packages/numpy/core/_methods.py:170: RuntimeWarning: invalid value encountered in true_divide
ret = ret.dtype.type(ret / rcount)

Any suggestion thank you in advance.

Hi,
The images used for the MRI case are not referenced to in the paper. Can you specify where are these images from?

Thank you

您好 我在训练MRI_plus版本时，遇到维度不匹配的问题，具体的报错如下： x = x - self.lambda_step * fft_forback(x, mask)
RuntimeError: The size of tensor a (2) must match the size of tensor b (256) at non-singleton dimension 5 请问这该如何解决呢？

Hi, I'm trying to train a model for MRI reconstruction but I am not able to make it work since it seems that the sampling matrices are not in the directory.

Could you please provide these matrices ? Thank you

您能分享对比算法中TVAL3 和D-AMP的代码么？谢谢

The size of the convolution kernel in the sampling process is 33*33,
Could this size be modified?

How is the sampling matrix obtained? which algorithm is used? Thanks a lot!