Architecture search concepts in Computer Vision - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is Neural Architecture Search (NAS)?

Neural Architecture Search is a method to automatically find the best design of a neural network for a specific task, instead of manually choosing the layers and connections.

beginner

Why is architecture search important in computer vision?

Because the right network design can improve accuracy and speed for tasks like image recognition, making models better at understanding pictures.

intermediate

What are the three main components of architecture search?

1. Search space: all possible network designs to explore. 2. Search strategy: how to explore these designs. 3. Performance estimation: how to quickly check if a design works well.

beginner

What is a search space in architecture search?

The search space is the set of all possible network structures that the search method can try, like different numbers of layers, types of layers, or connections.

intermediate

Name two common search strategies used in architecture search.

Two common strategies are: 1) Reinforcement Learning, where a controller learns to pick good designs, and 2) Evolutionary Algorithms, which mimic natural selection to improve designs over time.

What does the 'performance estimation' component do in architecture search?

A Measures how well a network design works quickly

B Defines all possible network designs

C Chooses which designs to try next

D Trains the final model on all data

Performance estimation quickly checks if a network design performs well, so the search can focus on promising designs.

Which of these is NOT typically part of the search space in architecture search?

A Connections between layers

B Type of activation functions

C Number of layers

D Color of the computer case

The color of the computer case is unrelated to neural network design and not part of the search space.

What is a benefit of using Neural Architecture Search?

A It guarantees 100% accuracy

B It makes networks run on any device without changes

C It removes the need for manual design of networks

D It replaces the need for training data

NAS automates the design process, reducing manual trial and error.

Which search strategy uses a controller to learn good network designs?

A Evolutionary Algorithms

B Reinforcement Learning

C Grid Search

D Random Search

Reinforcement Learning uses a controller that learns to pick better designs over time.

In architecture search, what does the term 'search strategy' mean?

A The way to explore different network designs

B The final model's accuracy

C The hardware used for training

D The dataset used for testing

Search strategy defines how the search process tries different network designs.

Explain the three main components of architecture search and why each is important.

Think about what you need to explore, how you explore, and how you check results.

Describe how Neural Architecture Search can improve computer vision models compared to manual design.

Consider the benefits of letting a computer find the best design.

Practice

1. What is the main goal of architecture search in computer vision models?

easy

A. To collect more training data

B. To manually tune model parameters

C. To automatically find the best model design

D. To reduce image resolution

Solution

1. Step 1: Understand architecture search purpose

   Architecture search aims to find the best model design automatically without manual trial and error.

2. Step 2: Compare options

   Options B, C, and D do not describe architecture search goals. Only To automatically find the best model design matches the goal.

3. Final Answer:

   To automatically find the best model design -> Option C

4. Quick Check:

   Architecture search = automatic best design [OK]

Hint: Architecture search = automatic model design finder [OK]

Common Mistakes:

• Confusing architecture search with data collection
• Thinking it manually tunes parameters
• Mixing it with image preprocessing

2. Which of the following is a correct way to describe a search space in architecture search?

easy

A. A set of possible model designs to explore

B. The training dataset used for the model

C. The final accuracy metric after training

D. The hardware used to run the model

Solution

1. Step 1: Define search space

   Search space is the collection of all possible model designs or configurations that the search will try.

2. Step 2: Eliminate incorrect options

   Options B, C, and D relate to data, metrics, or hardware, not the search space itself.

3. Final Answer:

   A set of possible model designs to explore -> Option A

4. Quick Check:

   Search space = possible designs [OK]

Hint: Search space = all model options to try [OK]

Common Mistakes:

• Confusing search space with dataset
• Thinking search space is a metric
• Mixing search space with hardware details

3. Consider this pseudocode for architecture search:

for model in search_space:
    accuracy = train_and_evaluate(model)
    if accuracy > best_accuracy:
        best_model = model
        best_accuracy = accuracy
print(best_accuracy)

What does this code output?

medium

A. The list of all models tested

B. The accuracy of the best model found

C. The training loss of the last model

D. The total number of models in search_space

Solution

1. Step 1: Analyze the loop

   The loop trains and evaluates each model, updating best_accuracy if current accuracy is higher.

2. Step 2: Understand the print statement

   After checking all models, it prints the highest accuracy found among them.

3. Final Answer:

   The accuracy of the best model found -> Option B

4. Quick Check:

   Prints best accuracy = highest accuracy [OK]

Hint: Code prints highest accuracy found during search [OK]

Common Mistakes:

• Thinking it prints number of models
• Confusing accuracy with loss
• Assuming it prints all models

4. The following code snippet is intended to find the best model architecture, but it has a bug:

best_accuracy = 0
for model in search_space:
    accuracy = train_and_evaluate(model)
    if accuracy < best_accuracy:
        best_model = model
        best_accuracy = accuracy
print(best_accuracy)

What is the bug?

medium

A. best_accuracy should start at 1 instead of 0

B. train_and_evaluate should return loss, not accuracy

C. The print statement should print best_model, not best_accuracy

D. The comparison operator should be > instead of <

Solution

1. Step 1: Understand the goal

   The goal is to find the model with the highest accuracy, so we want to update when accuracy is greater than best_accuracy.

2. Step 2: Identify the bug

   The code uses accuracy < best_accuracy, which updates for worse accuracy, so it should be accuracy > best_accuracy.

3. Final Answer:

   The comparison operator should be > instead of < -> Option D

4. Quick Check:

   Use > to find best accuracy [OK]

Hint: Best accuracy means use >, not < in comparison [OK]

Common Mistakes:

• Starting best_accuracy at wrong value
• Printing wrong variable
• Confusing accuracy with loss

5. You want to speed up architecture search by reducing the search space size. Which strategy is best?

hard

A. Limit model depth and number of layers to a smaller range

B. Increase the number of training epochs for each model

C. Use a slower but more accurate optimizer

D. Train all models on the full dataset without sampling

Solution

1. Step 1: Understand search space impact

   Reducing search space size means limiting the number of possible model designs to try.

2. Step 2: Evaluate options

   Limit model depth and number of layers to a smaller range reduces model complexity range, shrinking search space. Options A, B, and D increase training time or data size, slowing search.

3. Final Answer:

   Limit model depth and number of layers to a smaller range -> Option A

4. Quick Check:

   Smaller search space = fewer model options [OK]

Hint: Shrink search space by limiting model complexity [OK]

Common Mistakes:

• Thinking more training epochs speed up search
• Choosing slower optimizers to improve speed
• Using full dataset always speeds search