search

JuliaWolleb / Diffusion-based-Segmentation

This is the official Pytorch implementation of the paper "Diffusion Models for Implicit Image Segmentation Ensembles".
MIT License
278 stars 36 forks source link

test loss #38

Closed saisusmitha closed 11 months ago

saisusmitha commented 1 year ago

Hello, How to find test loss or validation loss?

JuliaWolleb commented 1 year ago

Hi

For validation, you can just plot it like the trainin loss, if you wish. However, during sampling, plotting the MSE loss does not make any sense, since we sample from Gaussian noise and go through the whole denoising process. What you can do is report the Dice score on the generated segmentation mask.