Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

About the input channels #19

Open
tonytonglt opened this issue Jul 7, 2020 · 1 comment
Open

About the input channels #19

tonytonglt opened this issue Jul 7, 2020 · 1 comment

Comments

@tonytonglt
Copy link

tonytonglt commented Jul 7, 2020

Hi.
As you mentioned in the paper, you changed the input channels from 3 to 9. However, when I look into your code, in models.py line 53, I found that num_channels = 3 + 6 + 2. Does it mean input channel of 11? What does 3, 6 and 2 indicate?
Looking forward to your reply. Thanks.

@99991
Copy link

99991 commented Jul 16, 2020

I think the channels are distributed like this:

  • 3 channels: This is the input image (group normalized).
  • 6 channels: Some kind of approximate alpha matte computed from the distance to the foreground and background of the trimap. See Distance Transform #17 for more details.
  • 2 channels: The foreground and background channels of the trimap.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants