A survey on Image Data Augmentation for Deep Learning
Abstract Deep convolutional neural networks have performed remarkably well on many Computer Vision tasks. However, these networks are heavily reliant on big data to avoid overfitting. Overfitting refers to the phenomenon when a network learns a function with very high variance such as to perfectly m...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2019-07-01
|
| Series: | Journal of Big Data |
| Subjects: | |
| Online Access: | http://link.springer.com/article/10.1186/s40537-019-0197-0 |
| _version_ | 1819064491954929664 |
|---|---|
| author | Connor Shorten Taghi M. Khoshgoftaar |
| author_facet | Connor Shorten Taghi M. Khoshgoftaar |
| author_sort | Connor Shorten |
| collection | DOAJ |
| description | Abstract Deep convolutional neural networks have performed remarkably well on many Computer Vision tasks. However, these networks are heavily reliant on big data to avoid overfitting. Overfitting refers to the phenomenon when a network learns a function with very high variance such as to perfectly model the training data. Unfortunately, many application domains do not have access to big data, such as medical image analysis. This survey focuses on Data Augmentation, a data-space solution to the problem of limited data. 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. 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. The application of augmentation methods based on GANs are heavily covered in this survey. 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. This survey will present existing methods for Data Augmentation, promising developments, and meta-level decisions for implementing Data Augmentation. 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. |
| first_indexed | 2024-12-21T15:31:25Z |
| format | Article |
| id | doaj.art-677e4d78a94b48aa897a5c41858b1562 |
| institution | Directory Open Access Journal |
| issn | 2196-1115 |
| language | English |
| last_indexed | 2024-12-21T15:31:25Z |
| publishDate | 2019-07-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | Journal of Big Data |
| spelling | doaj.art-677e4d78a94b48aa897a5c41858b15622022-12-21T18:58:45ZengSpringerOpenJournal of Big Data2196-11152019-07-016114810.1186/s40537-019-0197-0A survey on Image Data Augmentation for Deep LearningConnor Shorten0Taghi M. Khoshgoftaar1Department of Computer and Electrical Engineering and Computer Science, Florida Atlantic UniversityDepartment of Computer and Electrical Engineering and Computer Science, Florida Atlantic UniversityAbstract Deep convolutional neural networks have performed remarkably well on many Computer Vision tasks. However, these networks are heavily reliant on big data to avoid overfitting. Overfitting refers to the phenomenon when a network learns a function with very high variance such as to perfectly model the training data. Unfortunately, many application domains do not have access to big data, such as medical image analysis. This survey focuses on Data Augmentation, a data-space solution to the problem of limited data. 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. 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. The application of augmentation methods based on GANs are heavily covered in this survey. 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. This survey will present existing methods for Data Augmentation, promising developments, and meta-level decisions for implementing Data Augmentation. 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.http://link.springer.com/article/10.1186/s40537-019-0197-0Data AugmentationBig dataImage dataDeep LearningGANs |
| spellingShingle | Connor Shorten Taghi M. Khoshgoftaar A survey on Image Data Augmentation for Deep Learning Journal of Big Data Data Augmentation Big data Image data Deep Learning GANs |
| title | A survey on Image Data Augmentation for Deep Learning |
| title_full | A survey on Image Data Augmentation for Deep Learning |
| title_fullStr | A survey on Image Data Augmentation for Deep Learning |
| title_full_unstemmed | A survey on Image Data Augmentation for Deep Learning |
| title_short | A survey on Image Data Augmentation for Deep Learning |
| title_sort | survey on image data augmentation for deep learning |
| topic | Data Augmentation Big data Image data Deep Learning GANs |
| url | http://link.springer.com/article/10.1186/s40537-019-0197-0 |
| work_keys_str_mv | AT connorshorten asurveyonimagedataaugmentationfordeeplearning AT taghimkhoshgoftaar asurveyonimagedataaugmentationfordeeplearning AT connorshorten surveyonimagedataaugmentationfordeeplearning AT taghimkhoshgoftaar surveyonimagedataaugmentationfordeeplearning |