Data augmentation importance in Computer Vision - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is data augmentation in computer vision?

Data augmentation is a technique that creates new training images by modifying existing ones, like flipping, rotating, or changing colors, to help the model learn better.

beginner

Why is data augmentation important for training computer vision models?

It helps the model see more varied examples, which reduces overfitting and improves its ability to recognize objects in new, unseen images.

intermediate

How does data augmentation help with overfitting?

By increasing the diversity of training images, data augmentation prevents the model from memorizing exact images and encourages it to learn general patterns.

beginner

Name three common data augmentation techniques used in computer vision.

Common techniques include flipping images horizontally, rotating images by small angles, and adjusting brightness or contrast.

intermediate

Can data augmentation replace collecting more real data? Why or why not?

No, data augmentation helps but cannot fully replace real data because it only modifies existing images and may not capture all real-world variations.

What is the main goal of data augmentation in computer vision?

A Increase training data variety to improve model learning

B Reduce the size of the training dataset

C Make images look more colorful

D Remove noise from images

Data augmentation increases the variety of training data to help the model learn better and generalize well.

Which of the following is NOT a common data augmentation technique?

A Horizontal flipping

B Deleting pixels permanently

C Image rotation

D Adding random noise

Deleting pixels permanently is not a standard augmentation technique because it can remove important image information.

How does data augmentation affect overfitting?

A It increases overfitting

B It causes the model to memorize data

C It has no effect

D It reduces overfitting by adding variety

Data augmentation reduces overfitting by providing more varied examples for the model to learn from.

Why can't data augmentation fully replace collecting new real images?

A Because it only modifies existing images and may miss real-world variations

B Because augmented images are always blurry

C Because it is too slow

D Because it requires special hardware

Data augmentation modifies existing images but cannot create completely new real-world scenarios.

Which of these is a benefit of using data augmentation?

A Less memory usage

B Faster training time

C Improved model generalization

D Simpler model architecture

Data augmentation helps the model generalize better by exposing it to more varied data.

Explain why data augmentation is important for training computer vision models.

Think about how seeing more different images helps a model perform better on new images.

List and describe three common data augmentation techniques used in computer vision.

Consider simple ways to change images without changing their meaning.