mirrors
/
mmsegmentation
mirror of https://github.com/open-mmlab/mmsegmentation.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Spaces fixed.

pull/58/head
johnzja 2020-08-12 11:37:12 +08:00
parent 80fd7f6998
commit 887aec2027
1 changed files with 3 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
mmseg/models/backbones/fast_scnn.py
View File

@ -70,23 +70,18 @@ class LearningToDownsample(nn.Module):
class GlobalFeatureExtractor(nn.Module): class GlobalFeatureExtractor(nn.Module):
"""Global feature extractor module. """Global feature extractor module.
Args: Args:
in_channels (int): Number of input channels of the GFE module. in_channels (int): Number of input channels of the GFE module.
block_channels (tuple): Tuple of ints. Each int specifies the block_channels (tuple): Tuple of ints. Each int specifies the
number of output channels of each Inverted Residual module. number of output channels of each Inverted Residual module.
out_channels(int): Number of output channels of the GFE module. out_channels(int): Number of output channels of the GFE module.
expand_ratio (int): upsampling factor of each Inverted Residual
t (int): t parameter (upsampling factor) of each Inverted Residual
module. module.
num_blocks (tuple): Tuple of ints. Each int specifies the number of num_blocks (tuple): Tuple of ints. Each int specifies the number of
times each Inverted Residual module is repeated. times each Inverted Residual module is repeated.
pool_scales (tuple): Tuple of ints. Each int specifies the parameter pool_scales (tuple): Tuple of ints. Each int specifies the parameter
required in 'global average pooling' within PPM. required in 'global average pooling' within PPM.
conv_cfg (dict | None): Config of conv layers. Default: None conv_cfg (dict | None): Config of conv layers. Default: None
norm_cfg (dict | None): Config of norm layers. Default: norm_cfg (dict | None): Config of norm layers. Default:
dict(type='BN') dict(type='BN')
@ -158,19 +153,16 @@ class GlobalFeatureExtractor(nn.Module):
class FeatureFusionModule(nn.Module): class FeatureFusionModule(nn.Module):
"""Feature fusion module. """Feature fusion module.
Args: Args:
higher_in_channels (int): Number of input channels of the higher_in_channels (int): Number of input channels of the
higher-resolution branch. higher-resolution branch.
lower_in_channels (int): Number of input channels of the lower_in_channels (int): Number of input channels of the
lower-resolution branch. lower-resolution branch.
out_channels (int): Number of output channels. out_channels (int): Number of output channels.
scale_factor (int): Scale factor applied to the lower-res input. scale_factor (int): Scale factor applied to the lower-res input.
Should be coherent with the downsampling factor determined Should be coherent with the downsampling factor determined
by the GFE module. by the GFE module.
conv_cfg (dict | None): Config of conv layers. Default: None conv_cfg (dict | None): Config of conv layers. Default: None
norm_cfg (dict | None): Config of norm layers. Default: norm_cfg (dict | None): Config of norm layers. Default:
dict(type='BN') dict(type='BN')