This listing is intended to give readers a broader understanding of the context of Data Augmentation. The following few paragraphs will introduce other solutions available to avoid overfitting in Deep Learning models. The augmented data will represent a more comprehensive set of possible data points, thus minimizing the distance between the training and validation set, as well as any future testing sets.ĭata Augmentation, the focus of this survey, is not the only technique that has been developed to reduce overfitting. Data Augmentation is a very powerful method of achieving this. To build useful Deep Learning models, the validation error must continue to decrease with the training error. The graph below depicts what overfitting might look like when visualizing these accuracies over training epochs (Fig. One way to discover overfitting is to plot the training and validation accuracy at each epoch during training. Models with poor generalizability have overfitted the training data. Generalizability refers to the performance difference of a model when evaluated on previously seen data (training data) versus data it has never seen before (testing data). Improving the generalization ability of these models is one of the most difficult challenges. There are many branches of study that hope to improve current benchmarks by applying deep convolutional networks to Computer Vision tasks. The success of CNNs has spiked interest and optimism in applying Deep Learning to Computer Vision tasks. This series of convolutional transformations can create much lower-dimensional and more useful representations of images than what could possibly be hand-crafted. Convolutional layers sequentially downsample the spatial resolution of images while expanding the depth of their feature maps. These neural networks utilize parameterized, sparsely connected kernels which preserve the spatial characteristics of images. Deep neural networks have been successfully applied to Computer Vision tasks such as image classification, object detection, and image segmentation thanks to the development of convolutional neural networks (CNNs). This has been fueled by the advancement of deep network architectures, powerful computation, and access to big data. Readers will understand how Data Augmentation can improve the performance of their models and expand limited datasets to take advantage of the capabilities of big data.ĭeep Learning models have made incredible progress in discriminative tasks. This survey will present existing methods for Data Augmentation, promising developments, and meta-level decisions for implementing Data Augmentation. In addition to augmentation techniques, this paper will briefly discuss other characteristics of Data Augmentation such as test-time augmentation, resolution impact, final dataset size, and curriculum learning. The application of augmentation methods based on GANs are heavily covered in this survey.

The image augmentation algorithms discussed in this survey include geometric transformations, color space augmentations, kernel filters, mixing images, random erasing, feature space augmentation, adversarial training, generative adversarial networks, neural style transfer, and meta-learning. Data Augmentation encompasses a suite of techniques that enhance the size and quality of training datasets such that better Deep Learning models can be built using them. This survey focuses on Data Augmentation, a data-space solution to the problem of limited data. Unfortunately, many application domains do not have access to big data, such as medical image analysis. Overfitting refers to the phenomenon when a network learns a function with very high variance such as to perfectly model the training data. However, these networks are heavily reliant on big data to avoid overfitting. Deep convolutional neural networks have performed remarkably well on many Computer Vision tasks.

Image mixer 3 error disc 2