Why architecture design impacts performance in Computer Vision - Challenge Your Understanding

Challenge - 5 Problems

🧠 Conceptual

intermediate

How does increasing model depth affect performance?

In convolutional neural networks, what is the most common effect of increasing the number of layers (depth) on model performance?

A It reduces performance because deeper models cannot learn complex features.

B It always improves performance by capturing more complex features without any drawbacks.

C It can improve performance by learning complex features but may cause vanishing gradients and overfitting.

D It has no effect on performance since only the number of neurons matters.

💡 Hint

Think about what happens when a model becomes very deep and how training might be affected.

Increasing depth allows the model to learn more complex features, but very deep networks can suffer from vanishing gradients and overfitting, which can hurt performance.

❓ Predict Output

intermediate

Output shape after convolution and pooling layers

Given the following PyTorch model snippet, what is the output shape after the last layer?

Computer Vision

import torch
import torch.nn as nn

class SimpleCNN(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 16, kernel_size=3, padding=1)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(16, 32, kernel_size=3, padding=1)
    def forward(self, x):
        x = self.pool(torch.relu(self.conv1(x)))
        x = self.pool(torch.relu(self.conv2(x)))
        return x

model = SimpleCNN()
input_tensor = torch.randn(1, 3, 64, 64)
output = model(input_tensor)
output.shape

A torch.Size([1, 32, 8, 8])

B torch.Size([1, 16, 8, 8])

C torch.Size([1, 16, 16, 16])

D torch.Size([1, 32, 16, 16])

💡 Hint

Calculate the size after each convolution and pooling step.

Input 64x64 → conv1 keeps size 64x64 → pool halves to 32x32 → conv2 keeps size 32x32 → pool halves to 16x16. Channels after conv2 are 32.

❓ Hyperparameter

advanced

Choosing kernel size impact on feature extraction

How does increasing the convolution kernel size from 3x3 to 7x7 typically affect a CNN's ability to extract features?

A Larger kernels capture more global features but increase parameters and risk overfitting.

B Larger kernels always improve performance without any drawbacks.

C Larger kernels reduce the receptive field and limit feature extraction.

D Kernel size does not affect feature extraction, only the number of filters matters.

💡 Hint

Think about how kernel size relates to the area of the image the filter sees.

Larger kernels see more of the image at once, capturing broader features but increasing model size and overfitting risk.

❓ Metrics

advanced

Interpreting validation accuracy drop with deeper architecture

A CNN model with 10 layers achieves 85% validation accuracy. Increasing to 50 layers drops validation accuracy to 70%. What is the most likely reason?

A The dataset is too large for the deeper model to learn.

B The deeper model suffers from vanishing gradients and overfitting.

C The deeper model is underfitting due to too few parameters.

D Validation accuracy always decreases with more layers.

💡 Hint

Consider training difficulties with very deep networks.

Very deep networks can have vanishing gradients making training hard and can overfit if not regularized, causing validation accuracy to drop.

🔧 Debug

expert

Identifying cause of exploding gradients in deep CNN

Given a deep CNN training with exploding gradients, which architectural choice is the most likely cause?

A Using ReLU activation without batch normalization in a very deep network.

B Using batch normalization and residual connections in the network.

C Using small kernel sizes like 3x3 in all convolution layers.

D Using dropout layers after every convolution.

💡 Hint

Think about what helps stabilize training in deep networks.

ReLU without batch normalization or residual connections can cause exploding gradients in deep networks due to unstable activations.

Practice

1. Why does the design of a neural network architecture affect its performance on image tasks?

easy

A. Because it controls the size of the training dataset

B. Because it determines how well the model can learn important features from images

C. Because it decides the file format of the images

D. Because it changes the color of the images

Solution

1. Step 1: Understand the role of architecture in feature learning

   The architecture defines layers and connections that extract patterns from images.

2. Step 2: Connect architecture to model performance

   Better feature extraction leads to improved accuracy and generalization on tasks.

3. Final Answer:

   Because it determines how well the model can learn important features from images -> Option B

4. Quick Check:

   Architecture affects feature learning = D [OK]

Hint: Think about how model structure helps find image patterns [OK]

Common Mistakes:

• Confusing architecture with image properties
• Thinking architecture changes data format
• Believing architecture controls dataset size

2. Which of the following is the correct way to define a convolutional layer in a deep learning model using Python and PyTorch?

easy

A. nn.Conv2d(in_channels=3, out_channels=16, kernel_size=3, stride=1, padding=1)

B. nn.Linear(in_features=3, out_features=16)

C. nn.Conv1d(in_channels=3, out_channels=16, kernel_size=3)

D. nn.MaxPool2d(kernel_size=2, stride=2)

Solution

1. Step 1: Identify the convolutional layer syntax

   In PyTorch, Conv2d is used for 2D image convolutions with parameters for channels and kernel size.

2. Step 2: Check each option's layer type

   nn.Conv2d(in_channels=3, out_channels=16, kernel_size=3, stride=1, padding=1) correctly uses nn.Conv2d with proper parameters; others define different layers.

3. Final Answer:

   nn.Conv2d(in_channels=3, out_channels=16, kernel_size=3, stride=1, padding=1) -> Option A

4. Quick Check:

   Correct Conv2d syntax = B [OK]

Hint: Look for Conv2d with correct parameters for image layers [OK]

Common Mistakes:

• Confusing Conv2d with Linear or Conv1d layers
• Missing stride or padding parameters
• Choosing pooling layers instead of convolution

3. Consider this simplified CNN architecture for image classification:

model = nn.Sequential(
  nn.Conv2d(3, 8, 3, padding=1),
  nn.ReLU(),
  nn.MaxPool2d(2),
  nn.Conv2d(8, 16, 3, padding=1),
  nn.ReLU(),
  nn.MaxPool2d(2),
  nn.Flatten(),
  nn.Linear(16*8*8, 10)
)

If the input images are 32x32 pixels, what is the size of the feature map before flattening?

medium

A. 8 channels with 8x8 spatial size

B. 8 channels with 16x16 spatial size

C. 16 channels with 16x16 spatial size

D. 16 channels with 8x8 spatial size

Solution

1. Step 1: Calculate size after first Conv2d and MaxPool2d

   Input 32x32, Conv2d with padding=1 keeps size 32x32, MaxPool2d(2) halves to 16x16 with 8 channels.

2. Step 2: Calculate size after second Conv2d and MaxPool2d

   Conv2d keeps size 16x16 with 16 channels, MaxPool2d halves to 8x8 with 16 channels.

3. Final Answer:

   16 channels with 8x8 spatial size -> Option D

4. Quick Check:

   Pooling halves size twice = 8x8 with 16 channels [OK]

Hint: Each MaxPool2d(2) halves spatial size [OK]

Common Mistakes:

• Forgetting padding keeps size after convolution
• Not halving size after pooling
• Mixing channel counts with spatial dimensions

4. You have a CNN model that overfits training data but performs poorly on new images. Which architecture change can help reduce overfitting?

medium

A. Remove all pooling layers to keep more details

B. Increase the number of convolutional filters drastically

C. Add dropout layers to randomly ignore some neurons during training

D. Use a smaller batch size during training

Solution

1. Step 1: Understand overfitting and regularization

   Overfitting means the model memorizes training data; dropout helps by randomly ignoring neurons to generalize better.

2. Step 2: Evaluate options for reducing overfitting

   Adding dropout (A) is a common fix; increasing filters (B) may worsen overfitting; removing pooling (C) increases parameters; batch size (D) affects training stability but less impact on overfitting.

3. Final Answer:

   Add dropout layers to randomly ignore some neurons during training -> Option C

4. Quick Check:

   Dropout reduces overfitting = A [OK]

Hint: Use dropout to prevent memorizing training data [OK]

Common Mistakes:

• Thinking bigger models always reduce overfitting
• Removing pooling increases parameters and overfitting
• Confusing batch size effects with architecture changes

5. You want to design a model for real-time object detection on a mobile device. Which architectural choice best balances accuracy and speed?

hard

A. Use a lightweight architecture like MobileNet with depthwise separable convolutions

B. Use a very deep ResNet with 152 layers for highest accuracy

C. Use a fully connected network without convolutions

D. Use a large kernel size (e.g., 11x11) in all convolution layers

Solution

1. Step 1: Identify requirements for mobile real-time detection

   Mobile devices need fast, efficient models with good accuracy and low computation.

2. Step 2: Evaluate architectural options

   MobileNet uses depthwise separable convolutions to reduce computation while keeping accuracy; very deep ResNet is slow; fully connected networks lack spatial understanding; large kernels increase computation.

3. Final Answer:

   Use a lightweight architecture like MobileNet with depthwise separable convolutions -> Option A

4. Quick Check:

   MobileNet balances speed and accuracy = C [OK]

Hint: Choose lightweight models designed for mobile use [OK]

Common Mistakes:

• Picking very deep models ignoring speed constraints
• Using fully connected layers for images
• Choosing large kernels that slow down inference