Image augmentation transforms in Computer Vision - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to apply a horizontal flip to the image using torchvision transforms.

Computer Vision

transform = torchvision.transforms.Compose([torchvision.transforms.[1]()])

Drag options to blanks, or click blank then click option'

ARandomHorizontalFlip

BRandomVerticalFlip

CColorJitter

DRandomRotation

Attempts:

3 left

💡 Hint

Common Mistakes

Using RandomVerticalFlip instead flips the image upside down.

Using ColorJitter changes colors, not flips.

RandomRotation rotates the image, not flips.

The RandomHorizontalFlip transform flips the image horizontally, which is a common augmentation technique.

2fill in blank

medium

Complete the code to apply a random rotation of up to 30 degrees to the image.

Computer Vision

transform = torchvision.transforms.RandomRotation([1])

Drag options to blanks, or click blank then click option'

A15

B30

C45

D60

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing 15 rotates less than asked.

Choosing 45 or 60 rotates more than asked.

The argument 30 means the image will be randomly rotated between -30 and +30 degrees.

3fill in blank

hard

Fix the error in the code to correctly apply color jitter with brightness change.

Computer Vision

transform = torchvision.transforms.ColorJitter(brightness=[1])

Drag options to blanks, or click blank then click option'

A'0.5'

B50

C0.5

DTrue

Attempts:

3 left

💡 Hint

Common Mistakes

Using a string '0.5' causes a type error.

Using 50 is too large and invalid.

Using True is not a valid brightness value.

The brightness parameter expects a float like 0.5 to control brightness jitter.

4fill in blank

hard

Fill both blanks to create a transform that resizes images to 128x128 and then converts them to tensors.

Computer Vision

transform = torchvision.transforms.Compose([torchvision.transforms.Resize([1]), torchvision.transforms.[2]()])

Drag options to blanks, or click blank then click option'

A(128, 128)

BToTensor

CNormalize

DCenterCrop

Attempts:

3 left

💡 Hint

Common Mistakes

Using Normalize instead of ToTensor won't convert image format.

Using CenterCrop changes image size differently.

Passing a single int instead of tuple to Resize changes aspect ratio.

Resize expects a size tuple like (128, 128). To convert images to tensors, use ToTensor.

5fill in blank

hard

Fill all three blanks to create a transform pipeline that randomly crops 100x100 patches, applies horizontal flip, and normalizes with mean 0.5 and std 0.5.

Computer Vision

transform = torchvision.transforms.Compose([
    torchvision.transforms.RandomCrop([1]),
    torchvision.transforms.[2](),
    torchvision.transforms.Normalize(mean=[[3]], std=[0.5])
])

Drag options to blanks, or click blank then click option'

A(100, 100)

BRandomHorizontalFlip

C0.5

DRandomVerticalFlip

Attempts:

3 left

💡 Hint

Common Mistakes

Using RandomVerticalFlip flips vertically, not horizontally.

Using mean other than 0.5 changes normalization.

Passing single int instead of tuple to RandomCrop changes behavior.

RandomCrop needs the crop size tuple (100, 100). The horizontal flip transform is RandomHorizontalFlip. The mean for normalization is 0.5.