149 lines
6.1 KiB
Python
149 lines
6.1 KiB
Python
# Copyright (c) OpenMMLab. All rights reserved.
|
|
from mmcv.cnn import build_conv_layer, build_norm_layer
|
|
|
|
from mmpretrain.registry import MODELS
|
|
from .resnet import Bottleneck as _Bottleneck
|
|
from .resnet import ResLayer, ResNet
|
|
|
|
|
|
class Bottleneck(_Bottleneck):
|
|
"""Bottleneck block for ResNeXt.
|
|
|
|
Args:
|
|
in_channels (int): Input channels of this block.
|
|
out_channels (int): Output channels of this block.
|
|
groups (int): Groups of conv2.
|
|
width_per_group (int): Width per group of conv2. 64x4d indicates
|
|
``groups=64, width_per_group=4`` and 32x8d indicates
|
|
``groups=32, width_per_group=8``.
|
|
stride (int): stride of the block. Default: 1
|
|
dilation (int): dilation of convolution. Default: 1
|
|
downsample (nn.Module, optional): downsample operation on identity
|
|
branch. Default: None
|
|
style (str): `pytorch` or `caffe`. If set to "pytorch", the stride-two
|
|
layer is the 3x3 conv layer, otherwise the stride-two layer is
|
|
the first 1x1 conv layer.
|
|
conv_cfg (dict, optional): dictionary to construct and config conv
|
|
layer. Default: None
|
|
norm_cfg (dict): dictionary to construct and config norm layer.
|
|
Default: dict(type='BN')
|
|
with_cp (bool): Use checkpoint or not. Using checkpoint will save some
|
|
memory while slowing down the training speed.
|
|
"""
|
|
|
|
def __init__(self,
|
|
in_channels,
|
|
out_channels,
|
|
base_channels=64,
|
|
groups=32,
|
|
width_per_group=4,
|
|
**kwargs):
|
|
super(Bottleneck, self).__init__(in_channels, out_channels, **kwargs)
|
|
self.groups = groups
|
|
self.width_per_group = width_per_group
|
|
|
|
# For ResNet bottleneck, middle channels are determined by expansion
|
|
# and out_channels, but for ResNeXt bottleneck, it is determined by
|
|
# groups and width_per_group and the stage it is located in.
|
|
if groups != 1:
|
|
assert self.mid_channels % base_channels == 0
|
|
self.mid_channels = (
|
|
groups * width_per_group * self.mid_channels // base_channels)
|
|
|
|
self.norm1_name, norm1 = build_norm_layer(
|
|
self.norm_cfg, self.mid_channels, postfix=1)
|
|
self.norm2_name, norm2 = build_norm_layer(
|
|
self.norm_cfg, self.mid_channels, postfix=2)
|
|
self.norm3_name, norm3 = build_norm_layer(
|
|
self.norm_cfg, self.out_channels, postfix=3)
|
|
|
|
self.conv1 = build_conv_layer(
|
|
self.conv_cfg,
|
|
self.in_channels,
|
|
self.mid_channels,
|
|
kernel_size=1,
|
|
stride=self.conv1_stride,
|
|
bias=False)
|
|
self.add_module(self.norm1_name, norm1)
|
|
self.conv2 = build_conv_layer(
|
|
self.conv_cfg,
|
|
self.mid_channels,
|
|
self.mid_channels,
|
|
kernel_size=3,
|
|
stride=self.conv2_stride,
|
|
padding=self.dilation,
|
|
dilation=self.dilation,
|
|
groups=groups,
|
|
bias=False)
|
|
|
|
self.add_module(self.norm2_name, norm2)
|
|
self.conv3 = build_conv_layer(
|
|
self.conv_cfg,
|
|
self.mid_channels,
|
|
self.out_channels,
|
|
kernel_size=1,
|
|
bias=False)
|
|
self.add_module(self.norm3_name, norm3)
|
|
|
|
|
|
@MODELS.register_module()
|
|
class ResNeXt(ResNet):
|
|
"""ResNeXt backbone.
|
|
|
|
Please refer to the `paper <https://arxiv.org/abs/1611.05431>`__ for
|
|
details.
|
|
|
|
Args:
|
|
depth (int): Network depth, from {50, 101, 152}.
|
|
groups (int): Groups of conv2 in Bottleneck. Default: 32.
|
|
width_per_group (int): Width per group of conv2 in Bottleneck.
|
|
Default: 4.
|
|
in_channels (int): Number of input image channels. Default: 3.
|
|
stem_channels (int): Output channels of the stem layer. Default: 64.
|
|
num_stages (int): Stages of the network. Default: 4.
|
|
strides (Sequence[int]): Strides of the first block of each stage.
|
|
Default: ``(1, 2, 2, 2)``.
|
|
dilations (Sequence[int]): Dilation of each stage.
|
|
Default: ``(1, 1, 1, 1)``.
|
|
out_indices (Sequence[int]): Output from which stages. If only one
|
|
stage is specified, a single tensor (feature map) is returned,
|
|
otherwise multiple stages are specified, a tuple of tensors will
|
|
be returned. Default: ``(3, )``.
|
|
style (str): `pytorch` or `caffe`. If set to "pytorch", the stride-two
|
|
layer is the 3x3 conv layer, otherwise the stride-two layer is
|
|
the first 1x1 conv layer.
|
|
deep_stem (bool): Replace 7x7 conv in input stem with 3 3x3 conv.
|
|
Default: False.
|
|
avg_down (bool): Use AvgPool instead of stride conv when
|
|
downsampling in the bottleneck. Default: False.
|
|
frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
|
|
-1 means not freezing any parameters. Default: -1.
|
|
conv_cfg (dict | None): The config dict for conv layers. Default: None.
|
|
norm_cfg (dict): The config dict for norm layers.
|
|
norm_eval (bool): Whether to set norm layers to eval mode, namely,
|
|
freeze running stats (mean and var). Note: Effect on Batch Norm
|
|
and its variants only. Default: False.
|
|
with_cp (bool): Use checkpoint or not. Using checkpoint will save some
|
|
memory while slowing down the training speed. Default: False.
|
|
zero_init_residual (bool): Whether to use zero init for last norm layer
|
|
in resblocks to let them behave as identity. Default: True.
|
|
"""
|
|
|
|
arch_settings = {
|
|
50: (Bottleneck, (3, 4, 6, 3)),
|
|
101: (Bottleneck, (3, 4, 23, 3)),
|
|
152: (Bottleneck, (3, 8, 36, 3))
|
|
}
|
|
|
|
def __init__(self, depth, groups=32, width_per_group=4, **kwargs):
|
|
self.groups = groups
|
|
self.width_per_group = width_per_group
|
|
super(ResNeXt, self).__init__(depth, **kwargs)
|
|
|
|
def make_res_layer(self, **kwargs):
|
|
return ResLayer(
|
|
groups=self.groups,
|
|
width_per_group=self.width_per_group,
|
|
base_channels=self.base_channels,
|
|
**kwargs)
|