Replies: 3 comments
-
Hi! It's great to receive your feedback! We didn‘t use the independent shadow model, our shadow is decoupled from the background along with the cars. |
Beta Was this translation helpful? Give feedback.
0 replies
-
Actually, we think that it is the shared category-level object representation that helps the model to decompose shadow and ground plane appearances. The editing performance when we use instance-wise object representation indeed drops in shadow regions. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi, thanks for sharing your great work.
I noticed that the shadow looks realistic even when you move objects. I'm quite curious about it. Did you train the shadow with cars, or use an independent shadow model?
Beta Was this translation helpful? Give feedback.
All reactions