class ResNet(nn.Module):
"""ResNet backbone.
Args:
depth (int): Depth of resnet, from {18, 34, 50, 101, 152}.
pretrained (str | None): Name of pretrained model. Default: None.
in_channels (int): Channel num of input features. Default: 3.
num_stages (int): Resnet stages. Default: 4.
strides (Sequence[int]): Strides of the first block of each stage.
out_indices (Sequence[int]): Indices of output feature. Default: (3, ).
dilations (Sequence[int]): Dilation of each stage.
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. Default: ``pytorch``.
frozen_stages (int): Stages to be frozen (all param fixed). -1 means
not freezing any parameters. Default: -1.
conv_cfg (dict): Config for norm layers. Default: dict(type='Conv').
norm_cfg (dict):
Config for norm layers. required keys are `type` and
`requires_grad`. Default: dict(type='BN2d', requires_grad=True).
act_cfg (dict): Config for activate layers.
Default: dict(type='ReLU', inplace=True).
norm_eval (bool): Whether to set BN layers to eval mode, namely, freeze
running stats (mean and var). Default: False.
partial_bn (bool): Whether to use partial bn. Default: False.
with_cp (bool): Use checkpoint or not. Using checkpoint will save some
memory while slowing down the training speed. Default: False.
"""
arch_settings = {
18: (BasicBlock, (2, 2, 2, 2)),
34: (BasicBlock, (3, 4, 6, 3)),
50: (Bottleneck, (3, 4, 6, 3)),
101: (Bottleneck, (3, 4, 23, 3)),
152: (Bottleneck, (3, 8, 36, 3))
}
def __init__(self,
depth,
pretrained=None,
torchvision_pretrain=True,
in_channels=3,
num_stages=4,
out_indices=(3, ),
strides=(1, 2, 2, 2),
dilations=(1, 1, 1, 1),
style='pytorch',
frozen_stages=-1,
conv_cfg=dict(type='Conv'),
norm_cfg=dict(type='BN2d', requires_grad=True),
act_cfg=dict(type='ReLU', inplace=True),
norm_eval=False,
partial_bn=False,
with_cp=False):
super().__init__()
if depth not in self.arch_settings:
raise KeyError(f'invalid depth {depth} for resnet')
self.depth = depth
self.in_channels = in_channels
self.pretrained = pretrained
self.torchvision_pretrain = torchvision_pretrain
self.num_stages = num_stages
assert 1 <= num_stages <= 4
self.out_indices = out_indices
assert max(out_indices) < num_stages
self.strides = strides
self.dilations = dilations
assert len(strides) == len(dilations) == num_stages
self.style = style
self.frozen_stages = frozen_stages
self.conv_cfg = conv_cfg
self.norm_cfg = norm_cfg
self.act_cfg = act_cfg
self.norm_eval = norm_eval
self.partial_bn = partial_bn
self.with_cp = with_cp
self.block, stage_blocks = self.arch_settings[depth]
self.stage_blocks = stage_blocks[:num_stages]
self.inplanes = 64
self._make_stem_layer()
self.res_layers = []
for i, num_blocks in enumerate(self.stage_blocks):
stride = strides[i]
dilation = dilations[i]
planes = 64 * 2**i
res_layer = make_res_layer(self.block,
self.inplanes,
planes,
num_blocks,
stride=stride,
dilation=dilation,
style=self.style,
conv_cfg=conv_cfg,
norm_cfg=norm_cfg,
act_cfg=act_cfg,
with_cp=with_cp)
self.inplanes = planes * self.block.expansion
layer_name = f'layer{i + 1}'
self.add_module(layer_name, res_layer)
self.res_layers.append(layer_name)
self.feat_dim = self.block.expansion * 64 * 2**(
len(self.stage_blocks) - 1)
def _make_stem_layer(self):
"""Construct the stem layers consists of a conv+norm+act module and a
pooling layer."""
self.conv1 = ConvModule(self.in_channels,
64,
kernel_size=7,
stride=2,
padding=3,
bias=False,
conv_cfg=self.conv_cfg,
norm_cfg=self.norm_cfg,
act_cfg=self.act_cfg)
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
@staticmethod
def _load_conv_params(conv, state_dict_tv, module_name_tv,
loaded_param_names):
"""Load the conv parameters of resnet from torchvision.
Args:
conv (nn.Module): The destination conv module.
state_dict_tv (OrderedDict): The state dict of pretrained
torchvision model.
module_name_tv (str): The name of corresponding conv module in the
torchvision model.
loaded_param_names (list[str]): List of parameters that have been
loaded.
"""
weight_tv_name = module_name_tv + '.weight'
if conv.weight.data.shape == state_dict_tv[weight_tv_name].shape:
conv.weight.data.copy_(state_dict_tv[weight_tv_name])
loaded_param_names.append(weight_tv_name)
if getattr(conv, 'bias') is not None:
bias_tv_name = module_name_tv + '.bias'
if conv.bias.data.shape == state_dict_tv[bias_tv_name].shape:
conv.bias.data.copy_(state_dict_tv[bias_tv_name])
loaded_param_names.append(bias_tv_name)
@staticmethod
def _load_bn_params(bn, state_dict_tv, module_name_tv, loaded_param_names):
"""Load the bn parameters of resnet from torchvision.
Args:
bn (nn.Module): The destination bn module.
state_dict_tv (OrderedDict): The state dict of pretrained
torchvision model.
module_name_tv (str): The name of corresponding bn module in the
torchvision model.
loaded_param_names (list[str]): List of parameters that have been
loaded.
"""
for param_name, param in bn.named_parameters():
param_tv_name = f'{module_name_tv}.{param_name}'
param_tv = state_dict_tv[param_tv_name]
if param.data.shape == param_tv.shape:
param.data.copy_(param_tv)
loaded_param_names.append(param_tv_name)
for param_name, param in bn.named_buffers():
param_tv_name = f'{module_name_tv}.{param_name}'
# some buffers like num_batches_tracked may not exist
if param_tv_name in state_dict_tv:
param_tv = state_dict_tv[param_tv_name]
if param.data.shape == param_tv.shape:
param.data.copy_(param_tv)
loaded_param_names.append(param_tv_name)
def _load_torchvision_checkpoint(self, logger=None):
"""Initiate the parameters from torchvision pretrained checkpoint."""
state_dict_torchvision = _load_checkpoint(self.pretrained)
if 'state_dict' in state_dict_torchvision:
state_dict_torchvision = state_dict_torchvision['state_dict']
loaded_param_names = []
for name, module in self.named_modules():
if isinstance(module, ConvModule):
# we use a ConvModule to wrap conv+bn+relu layers, thus the
# name mapping is needed
if 'downsample' in name:
# layer{X}.{Y}.downsample.conv->layer{X}.{Y}.downsample.0
original_conv_name = name + '.0'
# layer{X}.{Y}.downsample.bn->layer{X}.{Y}.downsample.1
original_bn_name = name + '.1'
else:
# layer{X}.{Y}.conv{n}.conv->layer{X}.{Y}.conv{n}
original_conv_name = name
# layer{X}.{Y}.conv{n}.bn->layer{X}.{Y}.bn{n}
original_bn_name = name.replace('conv', 'bn')
self._load_conv_params(module.conv, state_dict_torchvision,
original_conv_name, loaded_param_names)
self._load_bn_params(module.bn, state_dict_torchvision,
original_bn_name, loaded_param_names)
# check if any parameters in the 2d checkpoint are not loaded
remaining_names = set(
state_dict_torchvision.keys()) - set(loaded_param_names)
if remaining_names:
logger.info(
f'These parameters in pretrained checkpoint are not loaded'
f': {remaining_names}')
def init_weights(self):
"""Initiate the parameters either from existing checkpoint or from
scratch."""
if isinstance(self.pretrained, str):
logger = get_root_logger()
if self.torchvision_pretrain:
# torchvision's
self._load_torchvision_checkpoint(logger)
else:
# ours
load_checkpoint(self,
self.pretrained,
strict=False,
logger=logger)
elif self.pretrained is None:
for m in self.modules():
if isinstance(m, nn.Conv2d):
kaiming_init(m)
elif isinstance(m, nn.BatchNorm2d):
constant_init(m, 1)
else:
raise TypeError('pretrained must be a str or None')
def forward(self, x):
"""Defines the computation performed at every call.
Args:
x (torch.Tensor): The input data.
Returns:
torch.Tensor: The feature of the input samples extracted
by the backbone.
"""
x = self.conv1(x)
x = self.maxpool(x)
outs = []
for i, layer_name in enumerate(self.res_layers):
res_layer = getattr(self, layer_name)
x = res_layer(x)
if i in self.out_indices:
outs.append(x)
if len(outs) == 1:
return outs[0]
return tuple(outs)
def _freeze_stages(self):
"""Prevent all the parameters from being optimized before
``self.frozen_stages``."""
if self.frozen_stages >= 0:
self.conv1.bn.eval()
for m in self.conv1.modules():
for param in m.parameters():
param.requires_grad = False
for i in range(1, self.frozen_stages + 1):
m = getattr(self, f'layer{i}')
m.eval()
for param in m.parameters():
param.requires_grad = False
def _partial_bn(self):
logger = get_root_logger()
logger.info('Freezing BatchNorm2D except the first one.')
count_bn = 0
for m in self.modules():
if isinstance(m, nn.BatchNorm2d):
count_bn += 1
if count_bn >= 2:
m.eval()
# shutdown update in frozen mode
m.weight.requires_grad = False
m.bias.requires_grad = False
def train(self, mode=True):
"""Set the optimization status when training."""
super().train(mode)
self._freeze_stages()
if mode and self.norm_eval:
for m in self.modules():
if isinstance(m, _BatchNorm):
m.eval()
if mode and self.partial_bn:
self._partial_bn()
class KpHead(nn.Module):
def __init__(self,
lateral_connect,
in_channels,
out_channels,
start_level,
fusion_level,
end_level,
csp_center_loss=None,
regr_h_loss=None,
regr_offset_loss=None,
conv_cfg=None,
norm_cfg=None,
act_cfg=None):
super(KpHead, self).__init__()
self.lateral_connect = lateral_connect
self.num_ins = len(in_channels)
self.start_level = start_level
self.fusion_level = fusion_level
self.end_level = end_level
self.concat_level = self.end_level - self.fusion_level
assert self.fusion_level >= self.start_level
if self.lateral_connect:
self.la_convs = nn.ModuleList()
self.deconvs = nn.ModuleList()
self.l2_norms = nn.ModuleList()
for i in range(self.concat_level):
if self.lateral_connect:
l_conv = ConvModule(
in_channels[i],
out_channels,
1,
conv_cfg=conv_cfg,
norm_cfg=norm_cfg,
act_cfg=act_cfg,
inplace=False)
self.la_convs.append(l_conv)
l2_norm = L2Normalization(out_channels, scale=10.0)
self.l2_norms.append(l2_norm)
for i in range(self.fusion_level + 1, self.end_level + 1):
if i == self.fusion_level + 1:
stride, padding = 2, 1
else:
stride, padding = 4, 0
d_conv = DeconvModule(in_channels[i],
out_channels,
kernel_size=4,
stride=stride,
padding=padding)
self.deconvs.append(d_conv)
self.feat = ConvModule(out_channels*self.concat_level,
out_channels,
3,
stride=1,
padding=1,
norm_cfg=dict(type='BN', requires_grad=True),
inplace=False)
self.cls_conv = nn.Conv2d(out_channels, 1, 1)
self.reg_conv = nn.Conv2d(out_channels, 1, 1)
self.offset_conv = nn.Conv2d(out_channels, 2, 1)
self.init_weights()
self.csp_center_loss = CSPCenterLoss(**csp_center_loss)
self.regr_loss = RegrHLoss(**regr_h_loss)
self.offset_loss = RegrOffsetLoss(**regr_offset_loss)
def init_weights(self):
for m in self.feat.modules():
if isinstance(m, nn.Conv2d):
nn.init.xavier_normal_(m.weight)
nn.init.xavier_normal_(self.cls_conv.weight)
nn.init.xavier_normal_(self.reg_conv.weight)
nn.init.xavier_normal_(self.offset_conv.weight)
nn.init.constant_(self.cls_conv.bias, -math.log(0.99/0.01))
nn.init.constant_(self.reg_conv.bias, 0)
nn.init.constant_(self.offset_conv.bias, 0)
def forward(self, inputs):
assert len(inputs) == self.num_ins
deconv_outs = []
if self.lateral_connect:
la_features = [
la_conv(inputs[i])
for i, la_conv in enumerate(self.la_convs)
]
for i in range(self.fusion_level + 1, self.end_level + 1):
deconv_outs.append(self.deconvs[i](la_features[i]))
else:
for i in range(self.fusion_level + 1, self.end_level + 1):
deconv_outs.append(self.deconvs[i-1](inputs[i]))
norm_outs = []
for i in range(len(deconv_outs)):
norm_outs.append(self.l2_norms[i](deconv_outs[i]))
cat_out = torch.cat(norm_outs, dim=1)
cat_out = self.feat(cat_out)
x_class = self.cls_conv(cat_out)
x_regr = self.reg_conv(cat_out)
x_offset = self.offset_conv(cat_out)
return [x_class, x_regr, x_offset]
def loss(self,
preds,
seman_map,
scale_map,
offset_map):
cls_pred, regr_pred, offset_pred = preds[0], preds[1], preds[2]
loss_cls = self.csp_center_loss(cls_pred, seman_map)
loss_regr = self.regr_loss(regr_pred, scale_map)
loss_offset = self.offset_loss(offset_pred, offset_map)
return dict(loss_cls=loss_cls, loss_regr=loss_regr, loss_offset=loss_offset)
