Data transforms in PyTorch - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to apply a random horizontal flip to an image tensor using PyTorch transforms.

PyTorch

transform = torchvision.transforms.RandomHorizontalFlip(p=[1])

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

A0

B1.0

C0.5

D2

Attempts:

3 left

💡 Hint

Common Mistakes

Using 1.0 flips every image, which might be too much.

Using 0 means no flip happens.

The probability parameter p in RandomHorizontalFlip controls how often the flip happens. 0.5 means the image flips half the time.

2fill in blank

medium

Complete the code to convert a PIL image to a tensor using PyTorch transforms.

PyTorch

transform = torchvision.transforms.Compose([torchvision.transforms.ToTensor(), torchvision.transforms.Normalize(mean=[0.5,0.5,0.5], std=[0.5,0.5,0.5])])
image_tensor = transform([1])

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

Aimage_pil

Bimage_path

Cimage_tensor

Dimage_array

Attempts:

3 left

💡 Hint

Common Mistakes

Passing a file path string instead of an image object.

Passing an already converted tensor.

The ToTensor() transform expects a PIL Image or numpy array as input. Here, image_pil is the correct input.

3fill in blank

hard

Fix the error in the code to resize an image to 128x128 pixels using PyTorch transforms.

PyTorch

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

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

A128

B128, 128

C[128, 128]

D(128, 128)

Attempts:

3 left

💡 Hint

Common Mistakes

Passing a list instead of a tuple.

Passing two separate integers without parentheses.

The Resize transform expects a size as an int or a tuple. To resize to width and height, use a tuple like (128, 128).

4fill in blank

hard

Fill both blanks to create a transform that converts an image to grayscale and then to a tensor.

PyTorch

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

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

AGrayscale

BToTensor

CResize

DNormalize

Attempts:

3 left

💡 Hint

Common Mistakes

Using Resize or Normalize instead of Grayscale.

Switching the order of transforms.

The Grayscale() transform converts the image to grayscale. Then ToTensor() converts it to a tensor.

5fill in blank

hard

Fill all three blanks to create a transform pipeline that resizes an image to 64x64, converts it to a tensor, and normalizes it with mean 0.5 and std 0.5.

PyTorch

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

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

AResize

B(64, 64)

CToTensor

DGrayscale

Attempts:

3 left

💡 Hint

Common Mistakes

Using Grayscale instead of ToTensor.

Passing size as a single integer instead of a tuple.

First, Resize((64, 64)) resizes the image. Then ToTensor() converts it to a tensor. Finally, Normalize adjusts the values.