Data augmentation for supervised learning with generative adversarial networks

dc.contributor.advisor Chinmay Hegde
dc.contributor.advisor Jin Tian
dc.contributor.author Podduturi, Manaswi
dc.contributor.department Computer Science
dc.date 2018-08-11T03:59:13.000
dc.date.accessioned 2020-06-30T03:11:06Z
dc.date.available 2020-06-30T03:11:06Z
dc.date.copyright Tue May 01 00:00:00 UTC 2018
dc.date.embargo 2001-01-01
dc.date.issued 2018-01-01
dc.description.abstract <p>Deep learning is a powerful technology that is revolutionizing automation in many industries. Deep learning models have a numerous number of parameters and tend to over-fit very often. Data plays a major role to successfully avoid overfitting and to exploit recent advancements in deep learning. However, collecting reliable data is a major limiting factor in many industries. This problem is usually tackled by using a combination of data augmentation, dropout, transfer learning, and batch normalization methods. In this paper, we explore the problem of data augmentation and common techniques employed in the field of image classification. The most successful strategy is to use a combination of rotation, translation, scaling, shearing, and flipping transformations. We experimentally evaluate and compare the performance of different data augmentation methods, using a subset of CIFAR-10dataset. Finally, we propose a framework to leverage generative adversarial networks(GANs)which are known to produce photo-realistic images for augmenting data. In the past, different frameworks have been proposed to leverage GANs in unsupervised and semi-supervised learning. Labeling samples generated by GANs is a difficult problem. In this paper, we propose a framework to do this. We take advantage of data distribution learned by the generator to train a back propagation model that projects a real image of the known label onto latent space. The learned latent space variables of real images are perturbed randomly, fed to the generator to generate synthetic images of that particular label. Through experiments we discovered that while adding more real data always outperforms any data augmentation techniques, supplementing data using proposed framework act as a better regularizer than traditional methods and hence has better generalization capability.</p>
dc.format.mimetype application/pdf
dc.identifier archive/lib.dr.iastate.edu/etd/16439/
dc.identifier.articleid 7446
dc.identifier.contextkey 12331456
dc.identifier.doi https://doi.org/10.31274/etd-180810-6069
dc.identifier.s3bucket isulib-bepress-aws-west
dc.identifier.submissionpath etd/16439
dc.identifier.uri https://dr.lib.iastate.edu/handle/20.500.12876/30622
dc.language.iso en
dc.source.bitstream archive/lib.dr.iastate.edu/etd/16439/Podduturi_iastate_0097M_17331.pdf|||Fri Jan 14 21:00:23 UTC 2022
dc.subject.disciplines Computer Sciences
dc.title Data augmentation for supervised learning with generative adversarial networks
dc.type article
dc.type.genre thesis
dspace.entity.type Publication
relation.isOrgUnitOfPublication f7be4eb9-d1d0-4081-859b-b15cee251456
thesis.degree.discipline Computer Science
thesis.degree.level thesis
thesis.degree.name Master of Science
File
Original bundle
Now showing 1 - 1 of 1
Name:
Podduturi_iastate_0097M_17331.pdf
Size:
1.06 MB
Format:
Adobe Portable Document Format
Description: