Advertisement

Why Does Image Data Augmentation Work As A Regularizer in Deep Learning?

In this article, we will demonstrate why data augmentation is known as a regularization technique. How to apply data augmentation to our model and whether it is used as a preprocessing technique or post-processing techniques...? All these questions are answered in this demonstration.
image data augmentation

The problem with deep learning models is they need lots of data to train a model. There are two major problems while training deep learning models is overfitting and underfitting of the model. Those problems are solved by data augmentation is a regularization technique that makes slight modifications to the images and used to generate data.

In this article, we will demonstrate why data augmentation is known as a regularization technique. How to apply data augmentation to our model and whether it is used as a preprocessing technique or post-processing techniques…? All these questions are answered in the below demonstration.  

Topics that we will demonstrate in this article:-

  • Data augmentation as a regularizer and data generator.
  • Implementing Data augmentation techniques.

Data Augmentation As a Regularizer and Data Generator

The regularization is a technique used to reduce the overfitting in the model. unnecessarily. In dealing with deep learning models, too much learning is also bad for the model to make a prediction with unseen data. If we get good results in training data and poor results in unseen data (test data, validation data) then it is framed as an overfitting problem. So now using data augmentation, we perform few transformations to the data like flipping, cropping, adding noise to the data, etc. 

As you know, deep learning models are data hungry, if we are lacking data then by using data augmentation transformations of the image we can generate data. Data augmentation is a preprocessing technique because we only work on the data to train our model. In this technique, we generate new instances of images by cropping, flipping, zooming, shearing an original image. So, whenever the training lacks the image dataset, using augmentation, we can create thousands of images to train the model perfectly.

Implementing Data augmentation techniques using OpenCV

Flipping the image

In the below code snippet we are going to flip the image.

import cv2
from google.colab.patches import cv2_imshow
img = cv2.imread('/content/images (10).jpg')
print(type(img))
print(img.shape)
img_flip_ud = cv2.flip(img, 0)
cv2_imshow(img_flip_ud)
img_flip_lr = cv2.flip(img, 1)
cv2_imshow(img_flip_lr)
img_flip_ud_lr = cv2.flip(img, -1)
cv2_imshow(img_flip_ud_lr)
image data augmentation

As you can see in the above output we flipped the image from 1 image we generated 4 images.

Resize an image

In the below code snippet we are going to rotate the image.

image=cv2.imread('/content/images (10).jpg')
cv2_imshow(image)
print(image.shape)
h1=300
w1=300
dimension = (w1, h1) 
resized_image = cv2.resize(image, dimension, interpolation = cv2.INTER_AREA)
cv2_imshow(resized_image)

Cropping an Image

In the below code snippet we are going to crop the image.

image=cv2.imread('/content/images (10).jpg')
cv2_imshow(image)
cropped_image = image[50:170, 150:250]
cv2_imshow(cropped_image)

Implementing Data augmentation techniques using Keras 

In the below code snippet we are performing data augmentation techniques using Keras.

from keras.preprocessing.image import ImageDataGenerator 
from keras.preprocessing.image import array_to_img
from keras.preprocessing.image import img_to_array
from keras.preprocessing.image import load_img 
generated_data = ImageDataGenerator( 
        rotation_range = 40, 
        shear_range = 0.2, 
        zoom_range = 0.2, 
        horizontal_flip = True, 
        brightness_range = (0.5, 1.5)) 
img = load_img('/content/images (10).jpg')  
x = img_to_array(img) 
x = x.reshape((1, ) + x.shape)  
i = 0
for batch in datagen.flow(x, batch_size = 1, 
                          save_to_dir ='preview',  
                          save_prefix ='image', save_format ='jpeg'): 
    i += 1
    if i > 5: 
        Break

Original image

image data augmentation

Image after zooming

image data augmentation

Image after apply rotation

image data augmentation

Image after adding brightness

Conclusion

In the above demonstration, we demonstrated how data augmentation is used as a regularizer and to generate data. This is one of the best practices used by many programmers.

Download our Mobile App

Prudhvi varma
AI enthusiast, Currently working with Analytics India Magazine. I have experience of working with Machine learning, Deep learning real-time problems, Neural networks, structuring and machine learning projects. I am a Computer Vision researcher and I am Interested in solving real-time computer vision problems.

Subscribe to our newsletter

Join our editors every weekday evening as they steer you through the most significant news of the day.
Your newsletter subscriptions are subject to AIM Privacy Policy and Terms and Conditions.

Our Upcoming Events

15th June | Online

Building LLM powered applications using LangChain

17th June | Online

Mastering LangChain: A Hands-on Workshop for Building Generative AI Applications

Jun 23, 2023 | Bangalore

MachineCon 2023 India

26th June | Online

Accelerating inference for every workload with TensorRT

MachineCon 2023 USA

Jul 21, 2023 | New York

Cypher 2023

Oct 11-13, 2023 | Bangalore

3 Ways to Join our Community

Telegram group

Discover special offers, top stories, upcoming events, and more.

Discord Server

Stay Connected with a larger ecosystem of data science and ML Professionals

Subscribe to our Daily newsletter

Get our daily awesome stories & videos in your inbox
MOST POPULAR