InfraGAN: A GAN architecture to transfer visible images to infrared domain

Simple item page
buir.contributor.authorÖzkanoğlu, Mehmet Akif
buir.contributor.orcidÖzkanoğlu, Mehmet Akif|0000-0003-2581-9525
dc.citation.epage76en_US
dc.citation.spage69en_US
dc.citation.volumeNumber155en_US
dc.contributor.authorÖzkanoğlu, Mehmet Akif
dc.contributor.authorOzer, S.
dc.date.accessioned2023-02-15T10:49:35Z
dc.date.available2023-02-15T10:49:35Z
dc.date.issued2022-02-03
dc.departmentDepartment of Computer Engineeringen_US
dc.description.abstractUtilizing both visible and infrared (IR) images in various deep learning based computer vision tasks has been a recent trend. Consequently, datasets having both visible and IR image pairs are desired in many applications. However, while large image datasets taken at the visible spectrum can be found in many domains, large IR-based datasets are not easily available in many domains. The lack of IR counterparts of the available visible image datasets limits existing deep algorithms to perform on IR images effectively. In this paper, to overcome with that challenge, we introduce a generative adversarial network (GAN) based solution and generate the IR equivalent of a given visible image by training our deep network to learn the relation between visible and IR modalities. In our proposed GAN architecture (InfraGAN), we introduce using structural similarity as an additional loss function. Furthermore, in our discriminator, we do not only consider the entire image being fake or real but also each pixel being fake or real. We evaluate our comparative results on three different datasets and report the state of the art results over five metrics when compared to Pix2Pix and ThermalGAN architectures from the literature. We report up to +16% better performance in Structural Similarity Index Measure (SSIM) over Pix2Pix and +8% better performance over ThermalGAN for VEDAI dataset. Further gains on different metrics and on different datasets are also reported in our experiments section.en_US
dc.description.provenanceSubmitted by Ezgi Uğurlu (ezgi.ugurlu@bilkent.edu.tr) on 2023-02-15T10:49:35Z No. of bitstreams: 1 InfraGAN_A_GAN_architecture_to_transfer_visible_images_to_infrared_domain.pdf: 2184961 bytes, checksum: c3072faef2ef85015383b4e7b749f0b6 (MD5)en
dc.description.provenanceMade available in DSpace on 2023-02-15T10:49:35Z (GMT). No. of bitstreams: 1 InfraGAN_A_GAN_architecture_to_transfer_visible_images_to_infrared_domain.pdf: 2184961 bytes, checksum: c3072faef2ef85015383b4e7b749f0b6 (MD5) Previous issue date: 2022-02-03en
dc.embargo.release2024-02-03
dc.identifier.doi10.1016/j.patrec.2022.01.026en_US
dc.identifier.eissn1872-7344
dc.identifier.issn0167-8655
dc.identifier.urihttp://hdl.handle.net/11693/111329
dc.language.isoEnglishen_US
dc.publisherElsevier BV * North-Hollanden_US
dc.relation.isversionofhttps://doi.org/10.1016/j.patrec.2022.01.026en_US
dc.source.titlePattern Recognition Lettersen_US
dc.subjectDomain transferen_US
dc.subjectGANsen_US
dc.subjectInfrared image generationen_US
dc.titleInfraGAN: A GAN architecture to transfer visible images to infrared domainen_US
dc.typeArticleen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
InfraGAN_A_GAN_architecture_to_transfer_visible_images_to_infrared_domain.pdf
Size:
2.08 MB
Format:
Adobe Portable Document Format
Description:
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.69 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

Scholarly Publications - Computer Engineering