Multimodal Unsupervised Image-to-Image Translation (MUNIT)

Multimodal Unsupervised Image-to-Image Translation (MUNIT) 을 읽고 간략하게 정리한 글입니다.


Multimodal Unsupervised Image-to-Image Translation (MUNIT)

Keywords : GAN, Image-to-Image translation, style transfer

  • image representation = content code + style code
    • content code : domain invariant
    • style code : domain specific

스크린샷 2020-12-22 오전 12 14 37

  • Image Translation == Recombine its content code with a random style code
    • random style code : style space of the target domain

스크린샷 2020-12-22 오전 12 14 23

Auto-encoder Architecture

스크린샷 2020-12-22 오전 12 19 33