Skip to content

Padding in convolve2NN #2817

Description

@geq1t

Hi,

Is there any particular reason why I can't use padding of zero in convolve2NN() to get a valid convolution? Right now, the padding argument must not contain any non-zero elements, which feels overly restrictive to me.

Thanks for any insights you can give me.

Regards,
Gé Vissers

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions