Vision Transformer (ViT) in Computer Vision - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is the main idea behind the Vision Transformer (ViT)?

ViT splits an image into small patches and treats them like words in a sentence, then uses a Transformer model to learn from these patches for image recognition.

intermediate

How does ViT process images differently from traditional convolutional neural networks (CNNs)?

ViT does not use convolution layers; instead, it divides images into patches and applies self-attention mechanisms to capture relationships between patches.

intermediate

What role do positional embeddings play in Vision Transformer models?

Positional embeddings add information about the location of each image patch, helping the model understand the order and position of patches since Transformers do not have built-in spatial awareness.

advanced

Why is pretraining on large datasets important for Vision Transformers?

ViTs need large amounts of data to learn good representations because they lack the built-in inductive biases of CNNs, so pretraining on big datasets helps them perform well on smaller tasks.

beginner

What metric is commonly used to evaluate the performance of a Vision Transformer on image classification tasks?

Accuracy is commonly used, which measures the percentage of images correctly classified by the model.

What does a Vision Transformer (ViT) use to represent an image before processing?

A Small patches of the image treated like tokens

B Pixels directly as input vectors

C Convolutional filters applied to the whole image

D Histogram of gradients

ViT splits the image into small patches and treats each patch as a token for the Transformer.

Why are positional embeddings necessary in ViT models?

A To provide spatial location information of patches

B To reduce the size of the input

C To add color information to patches

D To increase the number of patches

Transformers do not know the order or position of tokens, so positional embeddings tell the model where each patch is located.

Which of the following is NOT a characteristic of Vision Transformers?

A Require large datasets for training

B Use of self-attention to capture relationships

C Built-in convolutional layers

D Divide images into patches

ViTs do not use convolutional layers; they rely on self-attention mechanisms.

What is a common evaluation metric for ViT on classification tasks?

A Perplexity

B Mean squared error

C BLEU score

D Accuracy

Accuracy measures how many images are correctly classified, which is standard for classification.

How does ViT handle the spatial structure of images?

A By using convolutional filters

B By using positional embeddings

C By flattening the image into a vector

D By ignoring spatial information

ViT uses positional embeddings to keep track of where each patch is located in the image.

Explain how Vision Transformer (ViT) processes an image from input to output.

Think about how ViT treats image patches like words in a sentence.

Describe the advantages and challenges of using Vision Transformers compared to CNNs.

Consider data needs and model design differences.

Practice

1. What is the main purpose of splitting an image into patches in a Vision Transformer (ViT)?

easy

A. To reduce the image size by cropping

B. To convert the image into smaller parts that the transformer can process as tokens

C. To apply convolution filters on each patch separately

D. To increase the image resolution for better detail

Solution

1. Step 1: Understand ViT input processing

   ViT splits images into fixed-size patches to treat each patch like a word token in language models.

2. Step 2: Purpose of patch splitting

   This allows the transformer to process image patches as a sequence, enabling attention mechanisms to learn relationships.

3. Final Answer:

   To convert the image into smaller parts that the transformer can process as tokens -> Option B

4. Quick Check:

   Image patches = tokens for transformer [OK]

Hint: Think of patches as words in a sentence for the transformer [OK]

Common Mistakes:

• Confusing patch splitting with image resizing
• Thinking patches are processed by convolution
• Assuming patches increase image resolution

2. Which of the following is the correct way to add a class token to the patch embeddings in ViT using Python-like pseudocode?

easy

A. patches = torch.cat([class_token, patches], dim=1)

B. patches = torch.cat([patches, class_token], dim=1)

C. patches = torch.cat([patches, class_token], dim=0)

D. patches = torch.cat([class_token, patches], dim=0)

Solution

1. Step 1: Understand tensor concatenation dimension

   Patch embeddings are sequences along dimension 1 (batch, seq, embed); class token must be prepended along this dimension.

2. Step 2: Correct concatenation syntax

   Using torch.cat with dim=1 adds class_token at the start of the sequence correctly.

3. Final Answer:

   patches = torch.cat([class_token, patches], dim=1) -> Option A

4. Quick Check:

   Class token prepended along sequence dim = patches = torch.cat([class_token, patches], dim=1) [OK]

Hint: Class token goes first, concat along sequence dimension (dim=1) [OK]

Common Mistakes:

• Concatenating along wrong dimension (dim=0)
• Appending class token at the end instead of start
• Mixing order of tensors in concat

3. Given the following simplified ViT patch embedding code, what is the shape of patch_embeddings after processing a batch of 8 images of size 32x32 with patch size 8 and embedding dimension 64?

patch_size = 8
embedding_dim = 64
batch_size = 8
image_size = 32
num_patches = (image_size // patch_size) ** 2
patch_embeddings = torch.randn(batch_size, num_patches, embedding_dim)

medium

A. (16, 8, 64)

B. (8, 64, 16)

C. (8, 8, 64)

D. (8, 16, 64)

Solution

1. Step 1: Calculate number of patches

   Number of patches = (32 / 8)^2 = 4^2 = 16 patches per image.

2. Step 2: Determine patch_embeddings shape

   Shape is (batch_size, num_patches, embedding_dim) = (8, 16, 64).

3. Final Answer:

   (8, 16, 64) -> Option D

4. Quick Check:

   Batch=8, patches=16, embed=64 [OK]

Hint: Calculate patches as (image/patch)^2, then batch x patches x embed [OK]

Common Mistakes:

• Mixing embedding dimension and patch count order
• Calculating patches incorrectly
• Confusing batch size with patch count

4. You have this ViT code snippet that throws an error:

class_token = torch.randn(1, 1, 64)
patches = torch.randn(8, 16, 64)
input_seq = torch.cat([class_token, patches], dim=1)

What is the cause of the error?

medium

A. Embedding dimensions do not match

B. Wrong concatenation dimension; should be dim=0

C. class_token shape should be (8, 1, 64) to match batch size

D. Dimension mismatch because class_token sequence size is 1 but patches sequence size is 16

Solution

1. Step 1: Check batch size compatibility

   class_token has batch size 1, patches have batch size 8; they must match for concatenation.

2. Step 2: Fix class_token shape

   class_token should be repeated or created with shape (8, 1, 64) to match patches batch size.

3. Final Answer:

   class_token shape should be (8, 1, 64) to match batch size -> Option C

4. Quick Check:

   Batch sizes must match for concat [OK]

Hint: Match batch sizes before concatenating tensors [OK]

Common Mistakes:

• Ignoring batch size mismatch
• Changing wrong concat dimension
• Assuming embedding dims cause error

5. In a Vision Transformer model, why is the class token important for image classification tasks?

hard

A. It aggregates information from all patches via attention to produce a final image representation

B. It stores the positional information of patches

C. It applies convolution to patches before transformer layers

D. It normalizes the patch embeddings before feeding to the transformer

Solution

1. Step 1: Understand class token role

   The class token is a special token that attends to all patch tokens and gathers their information.

2. Step 2: Use in classification

   After transformer layers, the class token embedding is used as the image's summary representation for classification.

3. Final Answer:

   It aggregates information from all patches via attention to produce a final image representation -> Option A

4. Quick Check:

   Class token = image summary for classification [OK]

Hint: Class token collects info from patches for final decision [OK]

Common Mistakes:

• Confusing class token with positional encoding
• Thinking class token applies convolution
• Assuming class token normalizes embeddings