def feasc50(num_classes=200, nparts=1, seflag=False):
"""Constructs a ResNet-50 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
if seflag:
rd = [16, 32, 64, 128]
model = ResNet(SEBottleneck, [3, 4, 6, 3], num_classes=num_classes, rd=rd, nparts=nparts, seflag=True)
else:
model = ResNet(ResBottleneck, [3, 4, 6, 3], num_classes=num_classes, nparts=nparts, seflag=False)
model.avgpool = nn.AdaptiveAvgPool2d(1)
return model
def __init__(self, class_num=62):
super(bnneck, self).__init__()
resnet = ResNet(BasicBlock, [2, 2, 2, 2])
self.base_model = nn.Sequential(
resnet.conv1,
resnet.bn1,
resnet.relu,
resnet.layer1,
resnet.layer2,
resnet.layer3,
resnet.layer4
)
self.maxpool = nn.AdaptiveMaxPool2d(1)
self.bnneck = nn.BatchNorm1d(256)
self.bnneck.bias.requires_grad_(False) # no shift
self.reduce_layer = nn.Conv2d(512, 256, 1)
# self.classifier = ClassBlock(512, 1024)
self.fc1 = nn.Sequential(
nn.Linear(256, class_num))
self.fc2 = nn.Sequential(
nn.Linear(256, class_num))
self.fc3 = nn.Sequential(
nn.Linear(256, class_num))
self.fc4 = nn.Sequential(
nn.Linear(256, class_num))
def __init__(self, class_num=62):
super(res18, self).__init__()
model_ft = ResNet(BasicBlock, [2, 2, 2, 2])
self.base_model = nn.Sequential(*list(model_ft.children())[:-3])
# attention schema
self.avgpool = nn.AdaptiveAvgPool2d(1)
self.maxpool = nn.AdaptiveMaxPool2d(1)
self.sign = nn.Sigmoid()
in_plances = 256
ratio = 8
self.a_fc1 = nn.Conv2d(in_plances,in_plances//ratio,1,bias=False)
self.a_relu = nn.ReLU()
self.a_fc2 = nn.Conv2d(in_plances//ratio, in_plances, 1, bias=False)
self.avg_pool = nn.AdaptiveAvgPool2d(1)
self.max_pool = nn.AdaptiveMaxPool2d(1)
self.reduce_layer = nn.Conv2d(512, 256, 1)
# self.classifier = ClassBlock(512, 1024)
self.fc1 = nn.Sequential(nn.Dropout(0.5),
nn.Linear(256, class_num))
self.fc2 = nn.Sequential(nn.Dropout(0.5),
nn.Linear(256, class_num))
self.fc3 = nn.Sequential(nn.Dropout(0.5),
nn.Linear(256, class_num))
self.fc4 = nn.Sequential(nn.Dropout(0.5),
nn.Linear(256, class_num))
def se_resnet34(num_classes):
"""Constructs a ResNet-34 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(PSEBasicBlock, [3, 4, 6, 3], num_classes=num_classes)
return model
def se_resnet152(num_classes):
"""Constructs a ResNet-152 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(PSEBottleneck, [3, 8, 36, 3], num_classes=num_classes)
return model
def resnet152(pretrained=False, **kwargs):
if pretrained:
model = ResNet(Bottleneck, [3, 8, 36, 3], **kwargs)
pretrained_state_dict = torch.load(
'./Authority/resnet152-b121ed2d.pth'
) # load_url函数根据model_urls字典下载或导入相应的预训练模型
now_state_dict = model.state_dict() # 返回model模块的字典
pretrained_state_dict.pop('fc.weight') # 排除全连接层的参数(全连接层返回分类个数)
pretrained_state_dict.pop('fc.bias')
now_state_dict.update(pretrained_state_dict)
model.load_state_dict(now_state_dict)
# 最后通过调用model的load_state_dict方法用预训练的模型参数来初始化你构建的网络结构，
# 这个方法就是PyTorch中通用的用一个模型的参数初始化另一个模型的层的操作。load_state_dict方法还有一个重要的参数是strict，
# 该参数默认是True，表示预训练模型的层和你的网络结构层严格对应相等（比如层名和维度）
return model
return ResNet(Bottleneck, [3, 8, 36, 3], **kwargs)
def se_resnet18(num_classes=1_000):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(SEBasicBlock, [2, 2, 2, 2], num_classes=num_classes)
model.avgpool = nn.AdaptiveAvgPool2d(1)
return model
def SE_ResNet101(num_classes=1_000):
"""Constructs a ResNet-101 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(SEBottleneck, [3, 4, 23, 3], num_classes=num_classes)
model.avgpool = nn.AdaptiveAvgPool2d(1)
return model
def se_resnet34(num_classes=5):
"""Constructs a ResNet-34 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(SEBasicBlock, [3, 4, 6, 3], num_classes=num_classes)
model.avgpool = nn.AdaptiveAvgPool2d(1)
return model
def se_resnet152(num_classes=5):
"""Constructs a ResNet-152 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(SEBottleneck, [3, 8, 36, 3], num_classes=num_classes)
model.avgpool = nn.AdaptiveAvgPool2d(1)
return model
def se_resnet34(pretrained, num_classes=1000):
"""Constructs a SE-ResNet-34 model."""
model = ResNet(SEBasicBlock, [3, 4, 6, 3], num_classes=num_classes)
model.avgpool = nn.AdaptiveAvgPool2d(1)
if pretrained:
state_dict = torch.load('/home/kg2/se_resnet34_best.pth')['state_dict']
model.load_state_dict(state_dict, strict=False)
return model
def feasc18(num_classes=200, nparts=1):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(SEBasicBlock, [2, 2, 2, 2], num_classes=num_classes, nparts=nparts)
model.avgpool = nn.AdaptiveAvgPool2d(1)
return model
def resnet18(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [2, 2, 2, 2], **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
return model
def resnet18_l05_w05(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [1, 1, 1, 1], scale_factor=2, **kwargs)
model.name = 'ResNet18(length=05 width=05)'
return model
def dse_resnet50(num_classes=1000):
"""Constructs a ResNet-50 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(DSEBottleneck, [3, 4, 6, 3], num_classes=num_classes)
model.avgpool = nn.AdaptiveAvgPool2d(1)
return model
def resnet36_l2_w2(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [4, 4, 4, 4], scale_factor=0.5, **kwargs)
model.name = 'ResNet18(length=2 width=2)'
return model
def _se_resnet(arch, block, layers, pretrained, progress, **kwargs):
# adapted from the _resnet function in torch vision
model = ResNet(block, layers, **kwargs)
model.avgpool = nn.AdaptiveAvgPool2d(1)
if pretrained:
state_dict = load_state_dict_from_url(model_urls[arch],
progress=progress)
model.load_state_dict(state_dict)
return model
def _resnet(arch, block, layers, variant=None, **kwargs):
if variant is None:
model = ResNet(block, layers, **kwargs)
elif variant == "C":
model = ResNet_C(block, layers, **kwargs)
elif variant == "D":
model = ResNet_D(block, layers, **kwargs)
elif variant == "PA":
model = ResNetV2_C(block, layers, **kwargs)
return model
def resnet18_nr3_234(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [2, 2, 2, 2], noskip_by_layer=[False, True, True, True], **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
model.name = 'ResNet18_NR3_234'
return model
def resnet18_ep01(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [2, 2, 2, 2], disable_early_downsampling=True, **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
model.name = 'ResNet18_EP01'
return model
def se_resnet50(num_classes=5, pretrained=False):
"""Constructs a ResNet-50 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(SEBottleneck, [3, 4, 6, 3], num_classes=num_classes)
model.avgpool = nn.AdaptiveAvgPool2d(1)
if pretrained:
model.load_state_dict(load_state_dict_from_url(
"https://github.com/moskomule/senet.pytorch/releases/download/archive/seresnet50-60a8950a85b2b.pkl"))
return model
def resnet18_dp00_dspl2(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [4, 4, None, None], disable_early_pooling=True, disable_early_downsampling=True, **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
model.name = 'ResNet18_dp00_DSPL2'
return model
def se_resnet50(num_classes=1000, pretrained=False):
"""Constructs a ResNet-50 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(SEBottleneck, [3, 4, 6, 3], num_classes=num_classes)
model.avgpool = nn.AdaptiveAvgPool2d(1)
if pretrained:
model.load_state_dict(model_zoo.load_url("https://www.dropbox.com/s/xpq8ne7rwa4kg4c/seresnet50-60a8950a85b2b.pkl"))
return model
def __init__(self, class_num=62):
super(senet, self).__init__()
self.model = ResNet(
SEBasicBlock, [2, 2, 2, 2], num_classes=class_num)
self.model.fc = nn.Linear(512, 256)
self.model.avgpool = nn.AdaptiveAvgPool2d((1,1))
self.drop = nn.Dropout(0.5)
self.fc1 = nn.Linear(256, class_num)
self.fc2 = nn.Linear(256, class_num)
self.fc3 = nn.Linear(256, class_num)
self.fc4 = nn.Linear(256, class_num)
def resnet152(pretrained=False, **kwargs):
"""Constructs a ResNet-152 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(Bottleneck, [3, 8, 36, 3], **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet152']))
model.name = 'ResNet152'
return model
def resnet18_thin(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [4, 4, 4, 4], scale_factor=2, **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
model.name = 'ResNet18_Thin'
return model
def resnet18_dspl6(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [1, 1, 1, 1, 2, 2, None, None], **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
model.name = 'ResNet18_DSPL6'
return model
def resnet18noskip_dspl3(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
kwargs["noskip"] = True
model = ResNet(BasicBlock, [2, 3, 3, None], **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
model.name = 'ResNet18NoSkip_DSPL3'
return model
def resnet34(pretrained=False, **kwargs):
"""Constructs a ResNet-34 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [3, 4, 6, 3], **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet34']))
model.name = 'ResNet34'
return model
def __init__(self, pretrained=False, out_features=2):
"""
"""
super().__init__()
self.net = ResNet(BasicBlock, [2, 2, 2, 2])
if pretrained:
self.net.load_state_dict(
model_zoo.load_url(model_urls['resnet_18']))
# change last layer
self.net.fc = nn.Linear(self.net.fc.in_features, out_features)
self.out_features = out_features
