315 lines
8.9 KiB
Python
315 lines
8.9 KiB
Python
from __future__ import division, absolute_import
|
|
import torch.utils.model_zoo as model_zoo
|
|
from torch import nn
|
|
from torch.nn import functional as F
|
|
|
|
__all__ = ['pcb_p6', 'pcb_p4']
|
|
|
|
model_urls = {
|
|
'resnet18': 'https://download.pytorch.org/models/resnet18-5c106cde.pth',
|
|
'resnet34': 'https://download.pytorch.org/models/resnet34-333f7ec4.pth',
|
|
'resnet50': 'https://download.pytorch.org/models/resnet50-19c8e357.pth',
|
|
'resnet101': 'https://download.pytorch.org/models/resnet101-5d3b4d8f.pth',
|
|
'resnet152': 'https://download.pytorch.org/models/resnet152-b121ed2d.pth',
|
|
}
|
|
|
|
|
|
def conv3x3(in_planes, out_planes, stride=1):
|
|
"""3x3 convolution with padding"""
|
|
return nn.Conv2d(
|
|
in_planes,
|
|
out_planes,
|
|
kernel_size=3,
|
|
stride=stride,
|
|
padding=1,
|
|
bias=False
|
|
)
|
|
|
|
|
|
class BasicBlock(nn.Module):
|
|
expansion = 1
|
|
|
|
def __init__(self, inplanes, planes, stride=1, downsample=None):
|
|
super(BasicBlock, self).__init__()
|
|
self.conv1 = conv3x3(inplanes, planes, stride)
|
|
self.bn1 = nn.BatchNorm2d(planes)
|
|
self.relu = nn.ReLU(inplace=True)
|
|
self.conv2 = conv3x3(planes, planes)
|
|
self.bn2 = nn.BatchNorm2d(planes)
|
|
self.downsample = downsample
|
|
self.stride = stride
|
|
|
|
def forward(self, x):
|
|
residual = x
|
|
|
|
out = self.conv1(x)
|
|
out = self.bn1(out)
|
|
out = self.relu(out)
|
|
|
|
out = self.conv2(out)
|
|
out = self.bn2(out)
|
|
|
|
if self.downsample is not None:
|
|
residual = self.downsample(x)
|
|
|
|
out += residual
|
|
out = self.relu(out)
|
|
|
|
return out
|
|
|
|
|
|
class Bottleneck(nn.Module):
|
|
expansion = 4
|
|
|
|
def __init__(self, inplanes, planes, stride=1, downsample=None):
|
|
super(Bottleneck, self).__init__()
|
|
self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
|
|
self.bn1 = nn.BatchNorm2d(planes)
|
|
self.conv2 = nn.Conv2d(
|
|
planes,
|
|
planes,
|
|
kernel_size=3,
|
|
stride=stride,
|
|
padding=1,
|
|
bias=False
|
|
)
|
|
self.bn2 = nn.BatchNorm2d(planes)
|
|
self.conv3 = nn.Conv2d(
|
|
planes, planes * self.expansion, kernel_size=1, bias=False
|
|
)
|
|
self.bn3 = nn.BatchNorm2d(planes * self.expansion)
|
|
self.relu = nn.ReLU(inplace=True)
|
|
self.downsample = downsample
|
|
self.stride = stride
|
|
|
|
def forward(self, x):
|
|
residual = x
|
|
|
|
out = self.conv1(x)
|
|
out = self.bn1(out)
|
|
out = self.relu(out)
|
|
|
|
out = self.conv2(out)
|
|
out = self.bn2(out)
|
|
out = self.relu(out)
|
|
|
|
out = self.conv3(out)
|
|
out = self.bn3(out)
|
|
|
|
if self.downsample is not None:
|
|
residual = self.downsample(x)
|
|
|
|
out += residual
|
|
out = self.relu(out)
|
|
|
|
return out
|
|
|
|
|
|
class DimReduceLayer(nn.Module):
|
|
|
|
def __init__(self, in_channels, out_channels, nonlinear):
|
|
super(DimReduceLayer, self).__init__()
|
|
layers = []
|
|
layers.append(
|
|
nn.Conv2d(
|
|
in_channels, out_channels, 1, stride=1, padding=0, bias=False
|
|
)
|
|
)
|
|
layers.append(nn.BatchNorm2d(out_channels))
|
|
|
|
if nonlinear == 'relu':
|
|
layers.append(nn.ReLU(inplace=True))
|
|
elif nonlinear == 'leakyrelu':
|
|
layers.append(nn.LeakyReLU(0.1))
|
|
|
|
self.layers = nn.Sequential(*layers)
|
|
|
|
def forward(self, x):
|
|
return self.layers(x)
|
|
|
|
|
|
class PCB(nn.Module):
|
|
"""Part-based Convolutional Baseline.
|
|
|
|
Reference:
|
|
Sun et al. Beyond Part Models: Person Retrieval with Refined
|
|
Part Pooling (and A Strong Convolutional Baseline). ECCV 2018.
|
|
|
|
Public keys:
|
|
- ``pcb_p4``: PCB with 4-part strips.
|
|
- ``pcb_p6``: PCB with 6-part strips.
|
|
"""
|
|
|
|
def __init__(
|
|
self,
|
|
num_classes,
|
|
loss,
|
|
block,
|
|
layers,
|
|
parts=6,
|
|
reduced_dim=256,
|
|
nonlinear='relu',
|
|
**kwargs
|
|
):
|
|
self.inplanes = 64
|
|
super(PCB, self).__init__()
|
|
self.loss = loss
|
|
self.parts = parts
|
|
self.feature_dim = 512 * block.expansion
|
|
|
|
# backbone network
|
|
self.conv1 = nn.Conv2d(
|
|
3, 64, kernel_size=7, stride=2, padding=3, bias=False
|
|
)
|
|
self.bn1 = nn.BatchNorm2d(64)
|
|
self.relu = nn.ReLU(inplace=True)
|
|
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
|
|
self.layer1 = self._make_layer(block, 64, layers[0])
|
|
self.layer2 = self._make_layer(block, 128, layers[1], stride=2)
|
|
self.layer3 = self._make_layer(block, 256, layers[2], stride=2)
|
|
self.layer4 = self._make_layer(block, 512, layers[3], stride=1)
|
|
|
|
# pcb layers
|
|
self.parts_avgpool = nn.AdaptiveAvgPool2d((self.parts, 1))
|
|
self.dropout = nn.Dropout(p=0.5)
|
|
self.conv5 = DimReduceLayer(
|
|
512 * block.expansion, reduced_dim, nonlinear=nonlinear
|
|
)
|
|
self.feature_dim = reduced_dim
|
|
self.classifier = nn.ModuleList(
|
|
[
|
|
nn.Linear(self.feature_dim, num_classes)
|
|
for _ in range(self.parts)
|
|
]
|
|
)
|
|
|
|
self._init_params()
|
|
|
|
def _make_layer(self, block, planes, blocks, stride=1):
|
|
downsample = None
|
|
if stride != 1 or self.inplanes != planes * block.expansion:
|
|
downsample = nn.Sequential(
|
|
nn.Conv2d(
|
|
self.inplanes,
|
|
planes * block.expansion,
|
|
kernel_size=1,
|
|
stride=stride,
|
|
bias=False
|
|
),
|
|
nn.BatchNorm2d(planes * block.expansion),
|
|
)
|
|
|
|
layers = []
|
|
layers.append(block(self.inplanes, planes, stride, downsample))
|
|
self.inplanes = planes * block.expansion
|
|
for i in range(1, blocks):
|
|
layers.append(block(self.inplanes, planes))
|
|
|
|
return nn.Sequential(*layers)
|
|
|
|
def _init_params(self):
|
|
for m in self.modules():
|
|
if isinstance(m, nn.Conv2d):
|
|
nn.init.kaiming_normal_(
|
|
m.weight, mode='fan_out', nonlinearity='relu'
|
|
)
|
|
if m.bias is not None:
|
|
nn.init.constant_(m.bias, 0)
|
|
elif isinstance(m, nn.BatchNorm2d):
|
|
nn.init.constant_(m.weight, 1)
|
|
nn.init.constant_(m.bias, 0)
|
|
elif isinstance(m, nn.BatchNorm1d):
|
|
nn.init.constant_(m.weight, 1)
|
|
nn.init.constant_(m.bias, 0)
|
|
elif isinstance(m, nn.Linear):
|
|
nn.init.normal_(m.weight, 0, 0.01)
|
|
if m.bias is not None:
|
|
nn.init.constant_(m.bias, 0)
|
|
|
|
def featuremaps(self, x):
|
|
x = self.conv1(x)
|
|
x = self.bn1(x)
|
|
x = self.relu(x)
|
|
x = self.maxpool(x)
|
|
x = self.layer1(x)
|
|
x = self.layer2(x)
|
|
x = self.layer3(x)
|
|
x = self.layer4(x)
|
|
return x
|
|
|
|
def forward(self, x):
|
|
f = self.featuremaps(x)
|
|
v_g = self.parts_avgpool(f)
|
|
|
|
if not self.training:
|
|
v_g = F.normalize(v_g, p=2, dim=1)
|
|
return v_g.view(v_g.size(0), -1)
|
|
|
|
v_g = self.dropout(v_g)
|
|
v_h = self.conv5(v_g)
|
|
|
|
y = []
|
|
for i in range(self.parts):
|
|
v_h_i = v_h[:, :, i, :]
|
|
v_h_i = v_h_i.view(v_h_i.size(0), -1)
|
|
y_i = self.classifier[i](v_h_i)
|
|
y.append(y_i)
|
|
|
|
if self.loss == 'softmax':
|
|
return y
|
|
elif self.loss == 'triplet':
|
|
v_g = F.normalize(v_g, p=2, dim=1)
|
|
return y, v_g.view(v_g.size(0), -1)
|
|
else:
|
|
raise KeyError('Unsupported loss: {}'.format(self.loss))
|
|
|
|
|
|
def init_pretrained_weights(model, model_url):
|
|
"""Initializes model with pretrained weights.
|
|
|
|
Layers that don't match with pretrained layers in name or size are kept unchanged.
|
|
"""
|
|
pretrain_dict = model_zoo.load_url(model_url)
|
|
model_dict = model.state_dict()
|
|
pretrain_dict = {
|
|
k: v
|
|
for k, v in pretrain_dict.items()
|
|
if k in model_dict and model_dict[k].size() == v.size()
|
|
}
|
|
model_dict.update(pretrain_dict)
|
|
model.load_state_dict(model_dict)
|
|
|
|
|
|
def pcb_p6(num_classes, loss='softmax', pretrained=True, **kwargs):
|
|
model = PCB(
|
|
num_classes=num_classes,
|
|
loss=loss,
|
|
block=Bottleneck,
|
|
layers=[3, 4, 6, 3],
|
|
last_stride=1,
|
|
parts=6,
|
|
reduced_dim=256,
|
|
nonlinear='relu',
|
|
**kwargs
|
|
)
|
|
if pretrained:
|
|
init_pretrained_weights(model, model_urls['resnet50'])
|
|
return model
|
|
|
|
|
|
def pcb_p4(num_classes, loss='softmax', pretrained=True, **kwargs):
|
|
model = PCB(
|
|
num_classes=num_classes,
|
|
loss=loss,
|
|
block=Bottleneck,
|
|
layers=[3, 4, 6, 3],
|
|
last_stride=1,
|
|
parts=4,
|
|
reduced_dim=256,
|
|
nonlinear='relu',
|
|
**kwargs
|
|
)
|
|
if pretrained:
|
|
init_pretrained_weights(model, model_urls['resnet50'])
|
|
return model
|