Implementation of the Image Generation Model for Fashion Item Multi Domain Transformation Based on StarGAN v2 


Vol. 47,  No. 11, pp. 1876-1883, Nov.  2022
10.7840/kics.2022.47.11.1876


PDF
  Abstract

Various GANs have been recently developed and are being used in wide application areas. In this paper, we intend to utilize the GAN model to implement a domain transformation system for multi-detailed item on fashion items. For the purpose, we experimented and implemented StarGAN v2, which is objectively verified model that can generate images converted to the desired domain style by specifying conditions. Models were created for each detailed item to minimize labeling of datasets and accurately derive transformation results according to domains. Using the pretrained image generation model, we derive a result of synthesizing styles extracted from one image into another image, and we present a plan to allow multiple domains to be transformed for one image by utilizing all of the models. Similarity between real and generated images is measured for performance evaluation of each domain transformed by image generation model.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Related Articles
  Cite this article

[IEEE Style]

H. Jang, B. Son, J. Lee, "Implementation of the Image Generation Model for Fashion Item Multi Domain Transformation Based on StarGAN v2," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 11, pp. 1876-1883, 2022. DOI: 10.7840/kics.2022.47.11.1876.

[ACM Style]

Hyelim Jang, Bong-Ki Son, and Jaeho Lee. 2022. Implementation of the Image Generation Model for Fashion Item Multi Domain Transformation Based on StarGAN v2. The Journal of Korean Institute of Communications and Information Sciences, 47, 11, (2022), 1876-1883. DOI: 10.7840/kics.2022.47.11.1876.

[KICS Style]

Hyelim Jang, Bong-Ki Son, Jaeho Lee, "Implementation of the Image Generation Model for Fashion Item Multi Domain Transformation Based on StarGAN v2," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 11, pp. 1876-1883, 11. 2022. (https://doi.org/10.7840/kics.2022.47.11.1876)