{"id":2720,"date":"2019-08-22T06:15:06","date_gmt":"2019-08-22T06:15:06","guid":{"rendered":"https:\/\/blogs.mathworks.com\/deep-learning\/?p=2720"},"modified":"2021-04-06T15:49:43","modified_gmt":"2021-04-06T19:49:43","slug":"data-augmentation-for-image-classification-applications-using-deep-learning","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/deep-learning\/2019\/08\/22\/data-augmentation-for-image-classification-applications-using-deep-learning\/","title":{"rendered":"Data Augmentation for Image Classification Applications Using Deep Learning"},"content":{"rendered":"This post is from Oge Marques, PhD<\/a> and Professor of Engineering and Computer Science at FAU. Oge is an ACM Distinguished Speaker, book author<\/a>, and 2019-20 AAAS Leshner Fellow<\/a>. He also happens to be a MATLAB aficionado and has been using MATLAB in his classroom for more than 20 years. You can also follow him on Twitter (@ProfessorOge<\/a>) <\/span>\r\n\r\n\r\n
<\/h6>\r\nThe popularization of deep learning for image classification and many other computer vision tasks can be attributed, in part, to the availability of very large volumes of training data. The general consensus in the machine learning and deep learning community is that, all other things being equal, the more training data you have the better your model (and, consequently, its performance on the test set) will be.\r\n
<\/h6>\r\nThere are many cases, however, in which you might not have a large enough data set for training a model that can learn the desired classes in a way that will generalize well to new examples, i.e. will not overfit to the training set. This can often be due to the difficulty and cost<\/em> associated with the acquisition and labeling of images to build large training sets, whether cost is expressed in terms of dollars, human effort, computational resources, or the time consumed in the process.\r\n
<\/h6>\r\nIn this blog post we will focus on a technique called data augmentation<\/em>, which is used to augment the existing dataset in a way that is more cost-effective than further data collection. In the case of image classification applications, data augmentation is usually accomplished using simple geometric transformation techniques applied to the original images, such as cropping, rotating, resizing, translating, and flipping, which we'll discuss in more detail below.\r\n
<\/h6>\r\nThe effectiveness and benefits of data augmentation have been extensively documented in the literature: it has been shown that data augmentation can act as a regularizer in preventing overfitting in neural networks [1, 2] and improve performance in imbalanced class problems [3]. Moreover, the winning approaches to highly visible challenges and competitions in image classification (such as ImageNet) throughout the years have used data augmentation. This has also motivated recent work on the development of methods to allow a neural network to learn augmentations that best improve the classifier [4].\r\n
<\/h6>\r\n

Data Augmentation Implementation in MATLAB<\/strong><\/h2>\r\n
<\/h6>\r\nImage data augmentation can be achieved in two ways [5]:\r\n
<\/h6>\r\n
    \r\n \t
  1. offline augmentation<\/em>: which consists of performing the transformations to the images (potentially using MATLAB's batch image processing capabilities [6]) and saving the results on disk, thereby increasing the size of the dataset by a factor equal to the number of transformations performed. This can be acceptable for smaller datasets.<\/li>\r\n \t
  2. online augmentation<\/em> or augmentation on the fly<\/em>: which consists of performing transformations on the mini-batches that would be fed to the model during training. This method is preferred for larger datasets, to avoid a potentially explosive increase in storage requirements.<\/li>\r\n<\/ol>\r\n \r\n\r\nMATLAB provides an elegant and easy-to-use solution for online image data augmentation, which consists of two main components:\r\n