search

StanfordMIMI / DDM2

[ICLR2023] Official repository of DDM2: Self-Supervised Diffusion MRI Denoising with Generative Diffusion Models
140 stars 21 forks source link

2d-image #13

Closed Koid217 closed 1 year ago

Koid217 commented 1 year ago

Hello, I was wondering if this can be used for 2D images

tiangexiang commented 1 year ago

Theoretically yes! But you have to have multiple noisy observations of the same 2D image, which will give you a 3D volume of size: T (the number of observations) x W (width) x H (height).

Koid217 commented 1 year ago

Could T be 2 (for example, images with noise and images without noise)?

tiangexiang commented 1 year ago

Unfortunately, I don't think T=2 can work as expected. DDM2 roughly requires T>=5 to be functional. Also, in our experiments, we sifted out clean data and didn't use them for training so I don't think including the clean image is a good idea.

Koid217 commented 1 year ago

Thank you so much for your reply and advice！