module_7_html = """ Module 7: Deep Learning for Image Understanding

Module 7: Deep Learning for Image Understanding

1. Learning Objectives

Understand the basics of neural networks and convolutional layers
Learn about popular CNN architectures
Use PyTorch or TensorFlow to implement deep learning models

2. Neural Networks Recap

A neural network consists of layers of interconnected nodes ("neurons"). Each neuron performs a weighted sum of inputs followed by a non-linear activation function like ReLU or Sigmoid.

output = activation(Wx + b)

3. Convolutional Neural Networks (CNNs)

Convolution Layer: Applies filters to capture local patterns
Pooling Layer: Reduces dimensionality (e.g., max pooling)
Fully Connected Layer: Final classification layer

4. Famous Architectures

LeNet-5: Early CNN for digit recognition
AlexNet: First deep CNN to win ImageNet
VGG: Uses only 3x3 convolutions and depth
ResNet: Introduces residual connections

5. PyTorch Implementation Example


import torch.nn as nn

class SimpleCNN(nn.Module):
    def __init__(self):
        super(SimpleCNN, self).__init__()
        self.conv1 = nn.Conv2d(1, 32, kernel_size=3)
        self.pool = nn.MaxPool2d(2, 2)
        self.fc1 = nn.Linear(32 * 13 * 13, 10)

    def forward(self, x):
        x = self.pool(nn.functional.relu(self.conv1(x)))
        x = x.view(-1, 32 * 13 * 13)
        x = self.fc1(x)
        return x

6. Training Loop


for epoch in range(epochs):
    for inputs, labels in dataloader:
        optimizer.zero_grad()
        outputs = model(inputs)
        loss = criterion(outputs, labels)
        loss.backward()
        optimizer.step()

7. Transfer Learning

Use a pre-trained model like ResNet and fine-tune it for your dataset. Benefits:

Faster training
Better accuracy on small datasets

8. Assignment

Objective: Train a CNN from scratch and compare with transfer learning

1. Use MNIST or CIFAR-10 dataset
2. Build and train a small CNN
3. Fine-tune a pre-trained ResNet model
4. Compare test accuracy and training time

Due: End of Week 7