template_model.py

"""
InceptionV3 Network modified from https://github.com/pytorch/vision/blob/master/torchvision/models/inception.py
New changes: add softmax layer + option for freezing lower layers except fc
"""
import os

import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.utils.model_zoo as model_zoo
from torch.nn import Parameter

__all__ = ['MLP', 'BottleneckMLP', 'Inception3', 'inception_v3', 'End2EndModel']

model_urls = {
    # Downloaded inception model (optional)
    'downloaded': 'pretrained/inception_v3_google-1a9a5a14.pth',
    # Inception v3 ported from TensorFlow
    'inception_v3_google': 'https://download.pytorch.org/models/inception_v3_google-1a9a5a14.pth',
}


class End2EndModel(torch.nn.Module):
    def __init__(self, model1, model2, use_sigmoid=False):
        super(End2EndModel, self).__init__()
        self.first_model = model1
        self.sec_model = model2
        self.use_sigmoid = use_sigmoid

    def forward_stage2(self, stage1_out):
        if self.use_sigmoid:
            attr_outputs = [torch.nn.Sigmoid()(o) for o in stage1_out]
        else:
            attr_outputs = stage1_out

        stage2_inputs = attr_outputs
        stage2_inputs = torch.cat(stage2_inputs, dim=1)
        all_out = [self.sec_model(stage2_inputs)]
        all_out.extend(stage1_out)
        return all_out

    def forward(self, x):
        if self.first_model.training:
            outputs, aux_outputs = self.first_model(x)
            return self.forward_stage2(outputs), self.forward_stage2(aux_outputs)
        else:
            outputs = self.first_model(x)
            return self.forward_stage2(outputs)


class MLP(nn.Module):
    def __init__(self, input_dim, num_classes):
        super(MLP, self).__init__()
        self.linear = nn.Linear(input_dim, num_classes)  # softmax is automatically handled by loss function

    def forward(self, x):
        x = self.linear(x)
        return x


class BottleneckMLP(nn.Module):
    def __init__(self, n_attributes, input_size=784, encodings=False):
        super(BottleneckMLP, self).__init__()
        self.encodings = encodings
        self.n_attributes = n_attributes
        self.input_size = input_size

        self.fc1 = nn.Linear(input_size, 120)
        self.fc3 = nn.Linear(120, n_attributes)

    def forward(self, x):
        x = x.view(-1, self.input_size)
        x = F.relu(self.fc1(x))
        enc = x
        x = self.fc3(enc)
        # update structure of x to align with inception output
        out = []
        x = torch.swapaxes(x, 0, 1)
        for i in range(self.n_attributes):
            test = torch.unsqueeze(x[i], dim=1)
            out.append(test)

        if self.encodings:
            return out, enc
        else:
            return out


def inception_v3(pretrained, **kwargs):
    """Inception v3 model architecture from
    `"Rethinking the Inception Architecture for Computer Vision" <https://arxiv.org/abs/1512.00567>`_.

    .. note::
        **Important**: In contrast to the other models the inception_v3 expects tensors with a size of
        N x 3 x 299 x 299, so ensure your images are sized accordingly.

    Args:
        pretrained (bool): If True, returns a model pre-trained on ImageNet
        transform_input (bool): If True, preprocesses the input according to the method with which it
            was trained on ImageNet. Default: *False*
    """
    if pretrained:
        if 'transform_input' not in kwargs:
            kwargs['transform_input'] = True
        model = Inception3(**kwargs)
        if os.path.exists(model_urls.get('downloaded')):
            model.load_partial_state_dict(torch.load(model_urls['downloaded']))
        else:
            model.load_partial_state_dict(model_zoo.load_url(model_urls['inception_v3_google']))
        return model

    return Inception3(**kwargs)


class Inception3(nn.Module):

    def __init__(self, num_classes, aux_logits=True, transform_input=False, n_attributes=0, bottleneck=False,
                 encodings=False):
        """
        Args:
        num_classes: number of main task classes
        aux_logits: whether to also output auxiliary logits
        transform input: whether to invert the transformation by ImageNet (should be set to True later on)
        n_attributes: number of attributes to predict
        bottleneck: whether to make X -> C model
        """
        super(Inception3, self).__init__()
        self.aux_logits = aux_logits
        self.transform_input = transform_input
        self.n_attributes = n_attributes
        self.bottleneck = bottleneck
        self.encodings = encodings
        self.Conv2d_1a_3x3 = BasicConv2d(3, 32, kernel_size=3, stride=2)
        self.Conv2d_2a_3x3 = BasicConv2d(32, 32, kernel_size=3)
        self.Conv2d_2b_3x3 = BasicConv2d(32, 64, kernel_size=3, padding=1)
        self.Conv2d_3b_1x1 = BasicConv2d(64, 80, kernel_size=1)
        self.Conv2d_4a_3x3 = BasicConv2d(80, 192, kernel_size=3)
        self.Mixed_5b = InceptionA(192, pool_features=32)
        self.Mixed_5c = InceptionA(256, pool_features=64)
        self.Mixed_5d = InceptionA(288, pool_features=64)
        self.Mixed_6a = InceptionB(288)
        self.Mixed_6b = InceptionC(768, channels_7x7=128)
        self.Mixed_6c = InceptionC(768, channels_7x7=160)
        self.Mixed_6d = InceptionC(768, channels_7x7=160)
        self.Mixed_6e = InceptionC(768, channels_7x7=192)
        if aux_logits:
            self.AuxLogits = InceptionAux(768, num_classes, n_attributes=self.n_attributes, bottleneck=bottleneck)
        self.Mixed_7a = InceptionD(768)
        self.Mixed_7b = InceptionE(1280)
        self.Mixed_7c = InceptionE(2048)

        # separate fc layer for each prediction task. If main task is involved, it's always the first fc in the list
        self.all_fc = nn.ModuleList()

        if self.n_attributes > 0:
            if not bottleneck:  # multitasking
                self.all_fc.append(FC(2048, num_classes))
            for i in range(self.n_attributes):
                self.all_fc.append(FC(2048, 1))
        else:
            self.all_fc.append(FC(2048, num_classes))

        for m in self.modules():
            if isinstance(m, nn.Conv2d) or isinstance(m, nn.Linear):
                import scipy.stats as stats
                stddev = m.stddev if hasattr(m, 'stddev') else 0.1
                x = stats.truncnorm(-2, 2, scale=stddev)
                values = torch.as_tensor(x.rvs(m.weight.numel()), dtype=m.weight.dtype)
                values = values.view(m.weight.size())
                with torch.no_grad():
                    m.weight.copy_(values)
            elif isinstance(m, nn.BatchNorm2d):
                nn.init.constant_(m.weight, 1)
                nn.init.constant_(m.bias, 0)

    def forward(self, x):
        if self.transform_input:
            x_ch0 = torch.unsqueeze(x[:, 0], 1) * (0.229 / 0.5) + (0.485 - 0.5) / 0.5
            x_ch1 = torch.unsqueeze(x[:, 1], 1) * (0.224 / 0.5) + (0.456 - 0.5) / 0.5
            x_ch2 = torch.unsqueeze(x[:, 2], 1) * (0.225 / 0.5) + (0.406 - 0.5) / 0.5
            x = torch.cat((x_ch0, x_ch1, x_ch2), 1)
        # N x 3 x 299 x 299
        x = self.Conv2d_1a_3x3(x)
        # N x 32 x 149 x 149
        x = self.Conv2d_2a_3x3(x)
        # N x 32 x 147 x 147
        x = self.Conv2d_2b_3x3(x)
        # N x 64 x 147 x 147
        x = F.max_pool2d(x, kernel_size=3, stride=2)
        # N x 64 x 73 x 73
        x = self.Conv2d_3b_1x1(x)
        # N x 80 x 73 x 73
        x = self.Conv2d_4a_3x3(x)
        # N x 192 x 71 x 71
        x = F.max_pool2d(x, kernel_size=3, stride=2)
        # N x 192 x 35 x 35
        x = self.Mixed_5b(x)
        # N x 256 x 35 x 35
        x = self.Mixed_5c(x)
        # N x 288 x 35 x 35
        x = self.Mixed_5d(x)
        # N x 288 x 35 x 35
        x = self.Mixed_6a(x)
        # N x 768 x 17 x 17
        x = self.Mixed_6b(x)
        # N x 768 x 17 x 17
        x = self.Mixed_6c(x)
        # N x 768 x 17 x 17
        x = self.Mixed_6d(x)
        # N x 768 x 17 x 17
        x = self.Mixed_6e(x)
        # N x 768 x 17 x 17
        if self.training and self.aux_logits:
            out_aux = self.AuxLogits(x)
        # N x 768 x 17 x 17
        x = self.Mixed_7a(x)
        # N x 1280 x 8 x 8
        x = self.Mixed_7b(x)
        # N x 2048 x 8 x 8
        x = self.Mixed_7c(x)
        # N x 2048 x 8 x 8
        # Adaptive average pooling
        x = F.adaptive_avg_pool2d(x, (1, 1))
        # N x 2048 x 1 x 1
        x = F.dropout(x, training=self.training)
        # N x 2048 x 1 x 1
        x = x.view(x.size(0), -1)
        # N x 2048

        encoding_out = x

        out = []
        for fc in self.all_fc:
            out.append(fc(x))
        if self.training and self.aux_logits:
            return out, out_aux
        else:
            if self.encodings:
                return out, encoding_out
            else:
                return out

    def load_partial_state_dict(self, state_dict):
        """
        If dimensions of the current model doesn't match the pretrained one (esp for fc layer),
        load whichever weights that match
        """
        own_state = self.state_dict()
        for name, param in state_dict.items():
            if name not in own_state or 'fc' in name:
                continue
            if isinstance(param, Parameter):
                param = param.data
            own_state[name].copy_(param)


class FC(nn.Module):

    def __init__(self, input_dim, output_dim, stddev=None):
        """
        Extend standard Torch Linear layer to include the option of expanding into 2 Linear layers
        """
        super(FC, self).__init__()
        self.fc = nn.Linear(input_dim, output_dim)
        if stddev:
            self.fc.stddev = stddev

    def forward(self, x):
        x = self.fc(x)
        return x


class InceptionA(nn.Module):

    def __init__(self, in_channels, pool_features):
        super(InceptionA, self).__init__()
        self.branch1x1 = BasicConv2d(in_channels, 64, kernel_size=1)

        self.branch5x5_1 = BasicConv2d(in_channels, 48, kernel_size=1)
        self.branch5x5_2 = BasicConv2d(48, 64, kernel_size=5, padding=2)

        self.branch3x3dbl_1 = BasicConv2d(in_channels, 64, kernel_size=1)
        self.branch3x3dbl_2 = BasicConv2d(64, 96, kernel_size=3, padding=1)
        self.branch3x3dbl_3 = BasicConv2d(96, 96, kernel_size=3, padding=1)

        self.branch_pool = BasicConv2d(in_channels, pool_features, kernel_size=1)

    def forward(self, x):
        branch1x1 = self.branch1x1(x)

        branch5x5 = self.branch5x5_1(x)
        branch5x5 = self.branch5x5_2(branch5x5)

        branch3x3dbl = self.branch3x3dbl_1(x)
        branch3x3dbl = self.branch3x3dbl_2(branch3x3dbl)
        branch3x3dbl = self.branch3x3dbl_3(branch3x3dbl)

        branch_pool = F.avg_pool2d(x, kernel_size=3, stride=1, padding=1)
        branch_pool = self.branch_pool(branch_pool)

        outputs = [branch1x1, branch5x5, branch3x3dbl, branch_pool]
        return torch.cat(outputs, 1)


class InceptionB(nn.Module):

    def __init__(self, in_channels):
        super(InceptionB, self).__init__()
        self.branch3x3 = BasicConv2d(in_channels, 384, kernel_size=3, stride=2)

        self.branch3x3dbl_1 = BasicConv2d(in_channels, 64, kernel_size=1)
        self.branch3x3dbl_2 = BasicConv2d(64, 96, kernel_size=3, padding=1)
        self.branch3x3dbl_3 = BasicConv2d(96, 96, kernel_size=3, stride=2)

    def forward(self, x):
        branch3x3 = self.branch3x3(x)

        branch3x3dbl = self.branch3x3dbl_1(x)
        branch3x3dbl = self.branch3x3dbl_2(branch3x3dbl)
        branch3x3dbl = self.branch3x3dbl_3(branch3x3dbl)

        branch_pool = F.max_pool2d(x, kernel_size=3, stride=2)

        outputs = [branch3x3, branch3x3dbl, branch_pool]
        return torch.cat(outputs, 1)


class InceptionC(nn.Module):

    def __init__(self, in_channels, channels_7x7):
        super(InceptionC, self).__init__()
        self.branch1x1 = BasicConv2d(in_channels, 192, kernel_size=1)

        c7 = channels_7x7
        self.branch7x7_1 = BasicConv2d(in_channels, c7, kernel_size=1)
        self.branch7x7_2 = BasicConv2d(c7, c7, kernel_size=(1, 7), padding=(0, 3))
        self.branch7x7_3 = BasicConv2d(c7, 192, kernel_size=(7, 1), padding=(3, 0))

        self.branch7x7dbl_1 = BasicConv2d(in_channels, c7, kernel_size=1)
        self.branch7x7dbl_2 = BasicConv2d(c7, c7, kernel_size=(7, 1), padding=(3, 0))
        self.branch7x7dbl_3 = BasicConv2d(c7, c7, kernel_size=(1, 7), padding=(0, 3))
        self.branch7x7dbl_4 = BasicConv2d(c7, c7, kernel_size=(7, 1), padding=(3, 0))
        self.branch7x7dbl_5 = BasicConv2d(c7, 192, kernel_size=(1, 7), padding=(0, 3))

        self.branch_pool = BasicConv2d(in_channels, 192, kernel_size=1)

    def forward(self, x):
        branch1x1 = self.branch1x1(x)

        branch7x7 = self.branch7x7_1(x)
        branch7x7 = self.branch7x7_2(branch7x7)
        branch7x7 = self.branch7x7_3(branch7x7)

        branch7x7dbl = self.branch7x7dbl_1(x)
        branch7x7dbl = self.branch7x7dbl_2(branch7x7dbl)
        branch7x7dbl = self.branch7x7dbl_3(branch7x7dbl)
        branch7x7dbl = self.branch7x7dbl_4(branch7x7dbl)
        branch7x7dbl = self.branch7x7dbl_5(branch7x7dbl)

        branch_pool = F.avg_pool2d(x, kernel_size=3, stride=1, padding=1)
        branch_pool = self.branch_pool(branch_pool)

        outputs = [branch1x1, branch7x7, branch7x7dbl, branch_pool]
        return torch.cat(outputs, 1)


class InceptionD(nn.Module):

    def __init__(self, in_channels):
        super(InceptionD, self).__init__()
        self.branch3x3_1 = BasicConv2d(in_channels, 192, kernel_size=1)
        self.branch3x3_2 = BasicConv2d(192, 320, kernel_size=3, stride=2)

        self.branch7x7x3_1 = BasicConv2d(in_channels, 192, kernel_size=1)
        self.branch7x7x3_2 = BasicConv2d(192, 192, kernel_size=(1, 7), padding=(0, 3))
        self.branch7x7x3_3 = BasicConv2d(192, 192, kernel_size=(7, 1), padding=(3, 0))
        self.branch7x7x3_4 = BasicConv2d(192, 192, kernel_size=3, stride=2)

    def forward(self, x):
        branch3x3 = self.branch3x3_1(x)
        branch3x3 = self.branch3x3_2(branch3x3)

        branch7x7x3 = self.branch7x7x3_1(x)
        branch7x7x3 = self.branch7x7x3_2(branch7x7x3)
        branch7x7x3 = self.branch7x7x3_3(branch7x7x3)
        branch7x7x3 = self.branch7x7x3_4(branch7x7x3)

        branch_pool = F.max_pool2d(x, kernel_size=3, stride=2)
        outputs = [branch3x3, branch7x7x3, branch_pool]
        return torch.cat(outputs, 1)


class InceptionE(nn.Module):

    def __init__(self, in_channels):
        super(InceptionE, self).__init__()
        self.branch1x1 = BasicConv2d(in_channels, 320, kernel_size=1)

        self.branch3x3_1 = BasicConv2d(in_channels, 384, kernel_size=1)
        self.branch3x3_2a = BasicConv2d(384, 384, kernel_size=(1, 3), padding=(0, 1))
        self.branch3x3_2b = BasicConv2d(384, 384, kernel_size=(3, 1), padding=(1, 0))

        self.branch3x3dbl_1 = BasicConv2d(in_channels, 448, kernel_size=1)
        self.branch3x3dbl_2 = BasicConv2d(448, 384, kernel_size=3, padding=1)
        self.branch3x3dbl_3a = BasicConv2d(384, 384, kernel_size=(1, 3), padding=(0, 1))
        self.branch3x3dbl_3b = BasicConv2d(384, 384, kernel_size=(3, 1), padding=(1, 0))

        self.branch_pool = BasicConv2d(in_channels, 192, kernel_size=1)

    def forward(self, x):
        branch1x1 = self.branch1x1(x)

        branch3x3 = self.branch3x3_1(x)
        branch3x3 = [
            self.branch3x3_2a(branch3x3),
            self.branch3x3_2b(branch3x3),
        ]
        branch3x3 = torch.cat(branch3x3, 1)

        branch3x3dbl = self.branch3x3dbl_1(x)
        branch3x3dbl = self.branch3x3dbl_2(branch3x3dbl)
        branch3x3dbl = [
            self.branch3x3dbl_3a(branch3x3dbl),
            self.branch3x3dbl_3b(branch3x3dbl),
        ]
        branch3x3dbl = torch.cat(branch3x3dbl, 1)

        branch_pool = F.avg_pool2d(x, kernel_size=3, stride=1, padding=1)
        branch_pool = self.branch_pool(branch_pool)

        outputs = [branch1x1, branch3x3, branch3x3dbl, branch_pool]
        return torch.cat(outputs, 1)


class InceptionAux(nn.Module):

    def __init__(self, in_channels, num_classes, n_attributes=0, bottleneck=False):
        super(InceptionAux, self).__init__()
        self.conv0 = BasicConv2d(in_channels, 128, kernel_size=1)
        self.conv1 = BasicConv2d(128, 768, kernel_size=5)
        self.conv1.stddev = 0.01
        self.n_attributes = n_attributes
        self.bottleneck = bottleneck

        self.all_fc = nn.ModuleList()

        if n_attributes > 0:
            if not bottleneck:  # co training
                self.all_fc.append(FC(768, num_classes, stddev=0.001))
            for i in range(self.n_attributes):
                self.all_fc.append(FC(768, 1, stddev=0.001))
        else:
            self.all_fc.append(FC(768, num_classes, stddev=0.001))

    def forward(self, x):
        # N x 768 x 17 x 17
        x = F.avg_pool2d(x, kernel_size=5, stride=3)
        # N x 768 x 5 x 5
        x = self.conv0(x)
        # N x 128 x 5 x 5
        x = self.conv1(x)
        # N x 768 x 1 x 1
        # Adaptive average pooling
        x = F.adaptive_avg_pool2d(x, (1, 1))
        # N x 768 x 1 x 1
        x = x.view(x.size(0), -1)
        # N x 768
        out = []
        for fc in self.all_fc:
            out.append(fc(x))
        return out


class BasicConv2d(nn.Module):

    def __init__(self, in_channels, out_channels, **kwargs):
        super(BasicConv2d, self).__init__()
        self.conv = nn.Conv2d(in_channels, out_channels, bias=False, **kwargs)
        self.bn = nn.BatchNorm2d(out_channels, eps=0.001)

    def forward(self, x):
        x = self.conv(x)
        x = self.bn(x)
        return F.relu(x, inplace=True)