Image Data Augmentation techniques for Deep Learning -A Mirror Review

Making a computer understand the images is the task of computer vision. However, to familiarize it and learn to analyze what the computer is able to perceive is a tedious task. To address this challenge, Convolutional Neural Networks using deep learning frameworks could be performed and our goal could be achieved. For enhancing this task and making a computer learn more accurately and precisely about the variations in the real-world images, augmentation is an excellent way for making it possible. In this paper different types of normal augmentations, using deep learning methods for augmentation and other important works on this technique are discussed. This paper tries to cover the critical features of the augmentation, which will thus not only aid the new researchers but also other experienced researchers to interpret the latest trends going on in this technology, this would enable us to work together to make the computers learn more meticulously by utilizing various deep learning models that are trained on the images, making them more efficient and robust.