Abstract

Image-to-image translation involves translating images in one domain into images in another domain, while keeping some aspects of the image consistent across the domains. Image translation models that keep the category of the image consistent can be useful for applications like domain adaptation. Generative models like variational autoencoders have the ability to extract latent factors of generation from an image. Based on generative models like variational autoencoders and generative adversarial networks, we develop a semi-supervised image-to-image translation procedure. We apply this procedure to perform image translation and domain adaptation for complex digit datasets.

Original languageEnglish (US)
Title of host publicationSmart Multimedia - 1st International Conference, ICSM 2018, Revised Selected Papers
EditorsStefano Berretti, Anup Basu
PublisherSpringer Verlag
Pages334-344
Number of pages11
ISBN (Print)9783030043742
DOIs
StatePublished - Jan 1 2018
Event1st International Conference on Smart Multimedia, ICSM 2018 - Toulon, France
Duration: Aug 24 2018Aug 26 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11010 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other1st International Conference on Smart Multimedia, ICSM 2018
CountryFrance
CityToulon
Period8/24/188/26/18

Fingerprint

Generative Models
Digit
Model

Keywords

  • Domain adaptation
  • Generative adversarial networks
  • Image-to-image translation
  • Variational autoencoders

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Eusebio, J., Demakethepalli Venkateswara, H., & Panchanathan, S. (2018). Semi-supervised adversarial image-to-image translation. In S. Berretti, & A. Basu (Eds.), Smart Multimedia - 1st International Conference, ICSM 2018, Revised Selected Papers (pp. 334-344). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11010 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-04375-9_28

Semi-supervised adversarial image-to-image translation. / Eusebio, Jose; Demakethepalli Venkateswara, Hemanth; Panchanathan, Sethuraman.

Smart Multimedia - 1st International Conference, ICSM 2018, Revised Selected Papers. ed. / Stefano Berretti; Anup Basu. Springer Verlag, 2018. p. 334-344 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11010 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Eusebio, J, Demakethepalli Venkateswara, H & Panchanathan, S 2018, Semi-supervised adversarial image-to-image translation. in S Berretti & A Basu (eds), Smart Multimedia - 1st International Conference, ICSM 2018, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11010 LNCS, Springer Verlag, pp. 334-344, 1st International Conference on Smart Multimedia, ICSM 2018, Toulon, France, 8/24/18. https://doi.org/10.1007/978-3-030-04375-9_28
Eusebio J, Demakethepalli Venkateswara H, Panchanathan S. Semi-supervised adversarial image-to-image translation. In Berretti S, Basu A, editors, Smart Multimedia - 1st International Conference, ICSM 2018, Revised Selected Papers. Springer Verlag. 2018. p. 334-344. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-04375-9_28
Eusebio, Jose ; Demakethepalli Venkateswara, Hemanth ; Panchanathan, Sethuraman. / Semi-supervised adversarial image-to-image translation. Smart Multimedia - 1st International Conference, ICSM 2018, Revised Selected Papers. editor / Stefano Berretti ; Anup Basu. Springer Verlag, 2018. pp. 334-344 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{f98893fb770d44358ac67051dfc6b0a2,
title = "Semi-supervised adversarial image-to-image translation",
abstract = "Image-to-image translation involves translating images in one domain into images in another domain, while keeping some aspects of the image consistent across the domains. Image translation models that keep the category of the image consistent can be useful for applications like domain adaptation. Generative models like variational autoencoders have the ability to extract latent factors of generation from an image. Based on generative models like variational autoencoders and generative adversarial networks, we develop a semi-supervised image-to-image translation procedure. We apply this procedure to perform image translation and domain adaptation for complex digit datasets.",
keywords = "Domain adaptation, Generative adversarial networks, Image-to-image translation, Variational autoencoders",
author = "Jose Eusebio and {Demakethepalli Venkateswara}, Hemanth and Sethuraman Panchanathan",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-030-04375-9_28",
language = "English (US)",
isbn = "9783030043742",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "334--344",
editor = "Stefano Berretti and Anup Basu",
booktitle = "Smart Multimedia - 1st International Conference, ICSM 2018, Revised Selected Papers",

}

TY - GEN

T1 - Semi-supervised adversarial image-to-image translation

AU - Eusebio, Jose

AU - Demakethepalli Venkateswara, Hemanth

AU - Panchanathan, Sethuraman

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Image-to-image translation involves translating images in one domain into images in another domain, while keeping some aspects of the image consistent across the domains. Image translation models that keep the category of the image consistent can be useful for applications like domain adaptation. Generative models like variational autoencoders have the ability to extract latent factors of generation from an image. Based on generative models like variational autoencoders and generative adversarial networks, we develop a semi-supervised image-to-image translation procedure. We apply this procedure to perform image translation and domain adaptation for complex digit datasets.

AB - Image-to-image translation involves translating images in one domain into images in another domain, while keeping some aspects of the image consistent across the domains. Image translation models that keep the category of the image consistent can be useful for applications like domain adaptation. Generative models like variational autoencoders have the ability to extract latent factors of generation from an image. Based on generative models like variational autoencoders and generative adversarial networks, we develop a semi-supervised image-to-image translation procedure. We apply this procedure to perform image translation and domain adaptation for complex digit datasets.

KW - Domain adaptation

KW - Generative adversarial networks

KW - Image-to-image translation

KW - Variational autoencoders

UR - http://www.scopus.com/inward/record.url?scp=85058556095&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85058556095&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-04375-9_28

DO - 10.1007/978-3-030-04375-9_28

M3 - Conference contribution

AN - SCOPUS:85058556095

SN - 9783030043742

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 334

EP - 344

BT - Smart Multimedia - 1st International Conference, ICSM 2018, Revised Selected Papers

A2 - Berretti, Stefano

A2 - Basu, Anup

PB - Springer Verlag

ER -