FCN (Fully Convolutional Network) in Computer Vision - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a Fully Convolutional Network (FCN)?

An FCN is a type of neural network made only of convolutional layers. It can take images of any size and produce output maps of corresponding size, useful for tasks like image segmentation.

beginner

Why do FCNs not use fully connected layers?

FCNs replace fully connected layers with convolutional layers to keep spatial information and allow inputs of varying sizes, enabling pixel-wise predictions.

intermediate

How does an FCN produce pixel-wise predictions?

FCNs use convolutional layers to create feature maps and then upsample these maps (using methods like transposed convolution) to match the input image size, giving predictions for each pixel.

intermediate

What is the role of upsampling in FCNs?

Upsampling increases the spatial size of feature maps to the original image size, allowing the network to output detailed, pixel-level predictions for tasks like segmentation.

beginner

Name a common application of FCNs in computer vision.

FCNs are commonly used for semantic segmentation, where each pixel in an image is classified into a category, like labeling roads, cars, or people.

What type of layers does an FCN primarily use?

A Convolutional layers

B Fully connected layers

C Recurrent layers

D Pooling layers only

FCNs use convolutional layers instead of fully connected layers to maintain spatial information.

Why are fully connected layers avoided in FCNs?

A They are slower to train

B They increase model size unnecessarily

C They cause overfitting

D They reduce spatial information

Fully connected layers flatten spatial data, losing location information needed for pixel-wise tasks.

What is the purpose of upsampling in an FCN?

A To reduce the number of channels

B To increase spatial resolution of feature maps

C To add non-linearity

D To normalize the data

Upsampling restores the feature map size to match the input image for pixel-level predictions.

Which task is an FCN best suited for?

A Semantic segmentation

B Image classification

C Text generation

D Speech recognition

FCNs produce pixel-wise outputs, ideal for semantic segmentation.

How does an FCN handle input images of different sizes?

A It resizes all images to a fixed size

B It uses fully connected layers to adapt

C It processes images directly using convolutional layers

D It cannot handle different sizes

Convolutional layers allow FCNs to accept variable-sized inputs without resizing.

Explain how a Fully Convolutional Network (FCN) differs from a traditional CNN in terms of architecture and output.

Think about how FCNs keep spatial information and what that means for the output.

Describe the role of upsampling in FCNs and why it is important for tasks like semantic segmentation.

Consider what happens after convolution reduces image size.

Practice

1. What is the main advantage of a Fully Convolutional Network (FCN) compared to a traditional CNN for image tasks?

easy

A. It can label every pixel in the image, preserving spatial information.

B. It uses fully connected layers to improve classification accuracy.

C. It reduces the image size to a single vector for faster processing.

D. It only works with grayscale images for simplicity.

Solution

1. Step 1: Understand FCN structure

   FCNs replace fully connected layers with convolution layers to keep spatial info.

2. Step 2: Compare with traditional CNNs

   Traditional CNNs use fully connected layers that lose pixel location details, FCNs keep pixel-wise info.

3. Final Answer:

   It can label every pixel in the image, preserving spatial information. -> Option A

4. Quick Check:

   FCN labels pixels = B [OK]

Hint: FCNs keep pixel info; CNNs lose it with fully connected layers [OK]

Common Mistakes:

• Thinking FCNs use fully connected layers
• Assuming FCNs reduce image to vector
• Believing FCNs only work on grayscale

2. Which of the following is the correct way to replace a fully connected layer with a convolutional layer in an FCN?

easy

A. nn.Linear(in_features, out_features)

B. nn.ReLU()

C. nn.Conv2d(in_channels, out_channels, kernel_size=1)

D. nn.MaxPool2d(kernel_size=2)

Solution

1. Step 1: Identify layer types

   Fully connected layers are replaced by 1x1 convolution layers to keep spatial size.

2. Step 2: Match correct PyTorch layer

   nn.Conv2d with kernel size 1 acts like a fully connected layer per pixel.

3. Final Answer:

   nn.Conv2d(in_channels, out_channels, kernel_size=1) -> Option C

4. Quick Check:

   1x1 conv replaces FC layer = D [OK]

Hint: Use 1x1 conv to replace fully connected layers [OK]

Common Mistakes:

• Choosing nn.Linear which is fully connected
• Confusing pooling or activation layers as replacements
• Using wrong kernel sizes

3. Given this PyTorch FCN snippet, what is the output shape if the input is (1, 3, 64, 64)?

import torch
import torch.nn as nn

class SimpleFCN(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 16, 3, padding=1)
        self.conv2 = nn.Conv2d(16, 2, 1)
    def forward(self, x):
        x = self.conv1(x)
        x = self.conv2(x)
        return x

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

medium

A. (1, 2, 64, 64)

B. (1, 2, 62, 62)

C. (1, 16, 64, 64)

D. (1, 3, 64, 64)

Solution

1. Step 1: Analyze first conv layer

   Conv2d(3,16,3,padding=1) keeps spatial size 64x64, output shape (1,16,64,64).

2. Step 2: Analyze second conv layer

   Conv2d(16,2,1) with kernel size 1 keeps spatial size, output shape (1,2,64,64).

3. Final Answer:

   (1, 2, 64, 64) -> Option A

4. Quick Check:

   Padding keeps size; 1x1 conv keeps size = A [OK]

Hint: Padding keeps size; 1x1 conv keeps spatial dims [OK]

Common Mistakes:

• Ignoring padding effect on size
• Assuming kernel 3 reduces size without padding
• Confusing channel and spatial dimensions

4. You have an FCN model but the output size is smaller than the input image size. What is the most likely cause?

medium

A. Using 1x1 convolution layers increases spatial size.

B. Using convolution layers without padding reduces spatial size.

C. Using fully connected layers preserves spatial size.

D. Using batch normalization layers reduces spatial size.

Solution

1. Step 1: Understand convolution padding effect

   Convolution without padding shrinks spatial dimensions because kernel slides inside image.

2. Step 2: Identify cause of smaller output

   If output is smaller, likely padding is missing in conv layers causing shrinkage.

3. Final Answer:

   Using convolution layers without padding reduces spatial size. -> Option B

4. Quick Check:

   No padding shrinks output = A [OK]

Hint: No padding in conv layers shrinks output size [OK]

Common Mistakes:

• Thinking 1x1 conv changes size
• Believing batch norm affects size
• Assuming fully connected layers keep spatial size

5. You want to build an FCN for semantic segmentation on 128x128 images with 3 classes. Which architecture choice will produce output matching input size with class scores per pixel?

hard

A. Use fully connected layers after conv layers to classify whole image.

B. Use conv layers without padding and max pooling layers to reduce size.

C. Use conv layers with stride=2 and no upsampling layers.

D. Use conv layers with padding=1 and 1x1 conv at end, no pooling layers.

Solution

1. Step 1: Maintain spatial size

   Padding=1 with kernel size 3 keeps spatial size; no pooling avoids shrinking.

2. Step 2: Output class scores per pixel

   Use 1x1 conv with out_channels=3 to get class scores per pixel matching input size.

3. Final Answer:

   Use conv layers with padding=1 and 1x1 conv at end, no pooling layers. -> Option D

4. Quick Check:

   Padding + 1x1 conv = pixel-wise classes = C [OK]

Hint: Keep padding and no pooling for same size output [OK]

Common Mistakes:

• Using pooling reduces output size
• Using fully connected layers loses pixel info
• Using stride>1 without upsampling shrinks output