Python PyTorch – RandomHorizontalFlip() Function

In this article, we will discuss the RandomHorizontalFlip() Method in PyTorch Python.

RandomHorizontalFlip() method

RandomHorizontalFlip() method of torchvision.transforms module is used to horizontally flip the given image at a random angle with a given probability. This method accepts a PIL and tensor image as input. The tensor image is a PyTorch tensor with shape [C, H, W], where C represents the number of channels and  H, W represents the height and width respectively. This method returns a horizontally flipped image and an original image if the probability P is 1 or 0 respectively, if P is in the range between 0 to 1 then P is the probability to return the horizontally flipped image.

Syntax: torchvision.transforms.RandomHorizontalFlip(p)(img)

Parameter:

• p: p is the probability of the image being flipped at a random angle.
• img: input image to be flipped.  

Returns: This method returns a randomly flipped image  at a random angle.

The below image is used for demonstration:

Example 1:

In this example, we flip an image using RandomHorizontalFlip() Method when the probability is 1.

# import required libraries
import torch
import torchvision.transforms as T
from PIL import Image

# read input image from computer
img = Image.open('a.png')

# define a transform
transform = T.RandomHorizontalFlip(p=1)

# apply above defined transform to 
# input image
img = transform(img)

# display result
img.show()

Output:

Example 2:

In this example, we flip an image using RandomHorizontalFlip() Method when the probability is in the range of 0 to 1.

# import required libraries
import torch
import torchvision.transforms as T
from PIL import Image

# read input image from computer
img = Image.open('img.png')

# define a transform
transform = T.RandomHorizontalFlip(p=0.5)

# apply above defined transform to 
# input image
img = transform(img)

# display result
img.show()

Output: