Question about severity level (corruption) for depth estimation and semantic segmentation

mihirp1998 / Diffusion-TTA

Diffusion-TTA improves pre-trained discriminative models such as image classifiers or segmentors using pre-trained generative models.

https://diffusion-tta.github.io

Other

57 stars 4 forks source link

Question about severity level (corruption) for depth estimation and semantic segmentation #4

Open gabriel-tjio opened 9 months ago

gabriel-tjio commented 9 months ago

Thank you very much for uploading your code and for your paper. May I know what is the corruption severity level for both semantic segmentation and depth estimation? I tried looking for a description in the paper but it seems only image classification is specified (severity=5).

Thanks very much and I hope to hear from you.

mihirp1998 commented 9 months ago

Hi,

We consider severity=5 for both depth estimation and semantic segmentation.

gabriel-tjio commented 9 months ago

Thanks very much, Mihir!