search

JuliaWolleb / Diffusion-based-Segmentation

This is the official Pytorch implementation of the paper "Diffusion Models for Implicit Image Segmentation Ensembles".
MIT License
271 stars 35 forks source link

testing error #68

Open lanlinxxs opened 1 month ago

lanlinxxs commented 1 month ago

Hello, I encountered an error during prediction. Can you tell me why this happened? Traceback (most recent call last): File "/root/autodl-tmp/scripts/segmentation_sample.py", line 130, in main() File "/root/autodl-tmp/scripts/segmentation_sample.py", line 97, in main sample, x_noisy, org = sample_fn( File "/root/autodl-tmp/guided_diffusion/gaussian_diffusion.py", line 520, in p_sample_loop_known for sample in self.p_sample_loop_progressive( File "/root/autodl-tmp/guided_diffusion/gaussian_diffusion.py", line 586, in p_sample_loop_progressive out = self.p_sample( File "/root/autodl-tmp/guided_diffusion/gaussian_diffusion.py", line 433, in p_sample out = self.p_mean_variance( File "/root/autodl-tmp/guided_diffusion/respace.py", line 96, in p_mean_variance return super().p_mean_variance(self._wrap_model(model), *args, **kwargs) File "/root/autodl-tmp/guided_diffusion/gaussian_diffusion.py", line 316, in p_mean_variance self._predict_xstart_from_eps(x_t=x, t=t, eps=model_output) File "/root/autodl-tmp/guided_diffusion/gaussian_diffusion.py", line 337, in _predict_xstart_from_eps assert x_t.shape == eps.shape AssertionError

wildailurpodan commented 5 days ago

x_t shape: torch.Size([1, 1, 224, 224]) eps shape: torch.Size([1, 2, 224, 224]) i met the same error as you do. It looks like the expected output has two layers, is the original and the generated mask put together? Grayscale images should have only one layer