Skip to content

The final segmentation layer of the decoder in the code is inconsistent with the paper #72

@XianghaiKong

Description

@XianghaiKong

The final segmentation layer of the decoder in the code is as follows:
output = self.cls_seg(output)
The declaration of cls_seg in decode_head.py is as follows:
def cls_seg(self, feat):
"""Classify each pixel."""
if self.dropout is not None:
feat = self.dropout(feat)
output = self.conv_seg(feat)
return output
So is the final segmentation layer an MLP or a convolutional layer?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions