Hi, thank you for your great work!
I have a question about the performance of the pre trained model on SceneFlow.
When I crop the image size to 512x960 I get the epe error: 0.847
When I keep the image resolution and pad before test, I get the epe error: 0.905
How do you obtain the performance 0.7868 in Tab. 4(a)? Is it the same pre trained weight?
Hi, thank you for your great work! I have a question about the performance of the pre trained model on SceneFlow. When I crop the image size to 512x960 I get the epe error: 0.847 When I keep the image resolution and pad before test, I get the epe error: 0.905 How do you obtain the performance 0.7868 in Tab. 4(a)? Is it the same pre trained weight?