emotion_cnn_v2.py

"""
Enhanced CNN Architecture with Residual Connections and Attention
for Speech Emotion Recognition
"""

import torch
import torch.nn as nn
import torch.nn.functional as F

class ResidualBlock(nn.Module):
    """Residual block with batch normalization"""
    
    def __init__(self, in_channels, out_channels, stride=1):
        super(ResidualBlock, self).__init__()
        
        self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=3, 
                               stride=stride, padding=1, bias=False)
        self.bn1 = nn.BatchNorm2d(out_channels)
        self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3,
                               stride=1, padding=1, bias=False)
        self.bn2 = nn.BatchNorm2d(out_channels)
        
        # Shortcut connection
        self.shortcut = nn.Sequential()
        if stride != 1 or in_channels != out_channels:
            self.shortcut = nn.Sequential(
                nn.Conv2d(in_channels, out_channels, kernel_size=1,
                         stride=stride, bias=False),
                nn.BatchNorm2d(out_channels)
            )
    
    def forward(self, x):
        residual = x
        
        out = self.conv1(x)
        out = self.bn1(out)
        out = F.relu(out)
        
        out = self.conv2(out)
        out = self.bn2(out)
        
        out += self.shortcut(residual)
        out = F.relu(out)
        
        return out

class ChannelAttention(nn.Module):
    """Channel attention mechanism"""
    
    def __init__(self, channels, reduction=16):
        super(ChannelAttention, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        self.max_pool = nn.AdaptiveMaxPool2d(1)
        
        self.fc = nn.Sequential(
            nn.Linear(channels, channels // reduction, bias=False),
            nn.ReLU(),
            nn.Linear(channels // reduction, channels, bias=False)
        )
        self.sigmoid = nn.Sigmoid()
    
    def forward(self, x):
        b, c, _, _ = x.size()
        
        # Average pooling
        avg_out = self.fc(self.avg_pool(x).view(b, c))
        # Max pooling
        max_out = self.fc(self.max_pool(x).view(b, c))
        
        out = self.sigmoid(avg_out + max_out).view(b, c, 1, 1)
        return x * out.expand_as(x)

class ImprovedEmotionCNN(nn.Module):
    """
    Enhanced CNN with:
    - Residual connections
    - Channel attention
    - Deeper architecture
    - Better regularization
    """
    
    def __init__(self, num_classes=8, dropout_rate=0.4):
        super(ImprovedEmotionCNN, self).__init__()
        
        # Initial convolution
        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)
        
        # Residual blocks
        self.layer1 = self._make_layer(64, 64, 2, stride=1)
        self.layer2 = self._make_layer(64, 128, 2, stride=2)
        self.layer3 = self._make_layer(128, 256, 2, stride=2)
        self.layer4 = self._make_layer(256, 512, 2, stride=2)
        
        # Channel attention
        self.attention1 = ChannelAttention(64)
        self.attention2 = ChannelAttention(128)
        self.attention3 = ChannelAttention(256)
        self.attention4 = ChannelAttention(512)
        
        # Global pooling
        self.avgpool = nn.AdaptiveAvgPool2d((1, 1))
        self.maxpool_global = nn.AdaptiveMaxPool2d((1, 1))
        
        # Fully connected layers
        self.dropout = nn.Dropout(dropout_rate)
        self.fc1 = nn.Linear(512 * 2, 512)  # *2 because we concat avg and max pool
        self.bn_fc1 = nn.BatchNorm1d(512)
        self.fc2 = nn.Linear(512, 256)
        self.bn_fc2 = nn.BatchNorm1d(256)
        self.fc3 = nn.Linear(256, num_classes)
    
    def _make_layer(self, in_channels, out_channels, num_blocks, stride):
        layers = []
        layers.append(ResidualBlock(in_channels, out_channels, stride))
        for _ in range(1, num_blocks):
            layers.append(ResidualBlock(out_channels, out_channels, 1))
        return nn.Sequential(*layers)
    
    def forward(self, x):
        # Initial convolution
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)
        x = self.maxpool(x)
        
        # Residual blocks with attention
        x = self.layer1(x)
        x = self.attention1(x)
        
        x = self.layer2(x)
        x = self.attention2(x)
        
        x = self.layer3(x)
        x = self.attention3(x)
        
        x = self.layer4(x)
        x = self.attention4(x)
        
        # Global pooling (both avg and max)
        avg_pool = self.avgpool(x)
        max_pool = self.maxpool_global(x)
        x = torch.cat([avg_pool, max_pool], dim=1)
        x = x.view(x.size(0), -1)
        
        # Fully connected layers
        x = self.dropout(x)
        x = self.fc1(x)
        x = self.bn_fc1(x)
        x = F.relu(x)
        
        x = self.dropout(x)
        x = self.fc2(x)
        x = self.bn_fc2(x)
        x = F.relu(x)
        
        x = self.dropout(x)
        x = self.fc3(x)
        
        return x
    
    def get_num_params(self):
        """Calculate total number of parameters"""
        return sum(p.numel() for p in self.parameters() if p.requires_grad)

def test_model():
    """Test model architecture"""
    model = ImprovedEmotionCNN(num_classes=8)
    
    # Test input (updated for 196 features)
    batch_size = 4
    x = torch.randn(batch_size, 1, 196, 128)
    
    # Forward pass
    output = model(x)
    
    print("=" * 60)
    print("Enhanced Model Architecture Test")
    print("=" * 60)
    print(f"Input shape: {x.shape}")
    print(f"Output shape: {output.shape}")
    print(f"Total parameters: {model.get_num_params():,}")
    print("=" * 60)
    
    # Print model summary
    print("\nModel Summary:")
    print(model)

if __name__ == "__main__":
    test_model()