age_predictor_model.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Aug 19 12:17:34 2018

@author: kneehit
"""
import torch
import torch.nn as nn

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)



#%%
# Bottleneck block for ResNet to reduce dimensions
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
#%%
# Custom Convolution Neural Network architecture based on ResNet
class AgePredictor(nn.Module):
    
    # Define and Initialize Layers
    def __init__(self, block, layers, num_classes=1000):
        self.inplanes = 64
        super(AgePredictor, self).__init__()
        # ResNet Architecture
        self.conv1 = nn.Conv2d(1, 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=2)
        self.avgpool = nn.AdaptiveAvgPool2d(1)                          # <-       
        self.fc = nn.Linear(512 * block.expansion, 400)
        self.res_relu = nn.ReLU()
        
        
        # Fully Connected layer for gender
        self.gen_fc_1 = nn.Linear(1,16)
        self.gen_relu = nn.ReLU()
        
        # Feature Concatenation Layer
        self.cat_fc = nn.Linear(16+400,200)
        self.cat_relu = nn.ReLU()
        
        # Final Fully Connected Layer
        self.final_fc = nn.Linear(200,num_classes)
        
        # Simply using linear layer (w/o sigmoid) led to network predicting negative values for age
        # Therefore input was scaled to range from 0 and 1
        # and sigmoid is used as final layer to predict values which when 
        # denormalized led to positive values
        self.sigmoid = nn.Sigmoid()
        
        # Weight Initialization 
        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
            elif isinstance(m, nn.BatchNorm2d):
                nn.init.constant_(m.weight, 1)
                nn.init.constant_(m.bias, 0)

    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)

    # Forward Pass. x = Image tensor, y = gender tensor
    def forward(self, x,y):
# =============================================================================
#       ResNet Layers        
# =============================================================================
        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)

        x = self.avgpool(x)
        x = x.view(x.size(0), -1)
        x = self.fc(x)
        x = self.res_relu(x)
        x = x.view(x.size(0), -1)
        

# =============================================================================
#       Gender Fully Connected Layer
# =============================================================================
        y = self.gen_fc_1(y)
        y = self.gen_relu(y)
        y = y.view(y.size(0), -1)

        
# =============================================================================
#       Feature Concatenation Layer
# =============================================================================
      
        z = torch.cat((x,y),dim = 1)
        z = self.cat_fc(z)
        z = self.cat_relu(z)

# =============================================================================
#       Final FC Layer
# =============================================================================
        
        z = self.final_fc(z)
        z = self.sigmoid(z)

        return z