Results
Test results for the Supercat model are presented at unimelbmdap.github.io/supercat-results, which shows super-resolution outputs for two- and three-dimensional micro-CT images of rocks.
For example, a carbonate slice from the 2D test set upscaled by a factor of four with the WiDiT diffusion model:
Low-resolution input (125 × 125) |
Supercat prediction (500 × 500) |
|---|---|
|
|
The results use the DeepRock-SR 2D and 3D datasets introduced by Wang, Armstrong and Mostaghimi:
Wang, Y. D., Armstrong, R. T. & Mostaghimi, P. (2020). Boosting Resolution and Recovering Texture of 2D and 3D Micro-CT Images with Deep Learning. Water Resources Research, 56(1), e2019WR026052.
Wang, Y. D., Armstrong, R. & Mostaghimi, P. (2019). A Diverse Super Resolution Dataset of Digital Rocks (DeepRock-SR): Sandstone, Carbonate, and Coal. Digital Rocks Portal.