Contrastive Learning for Unpaired Image-to-Image Translation

Поделиться
HTML-код
  • Опубликовано: 22 дек 2024

Комментарии • 14

  • @DeepGamingAI
    @DeepGamingAI 4 года назад +5

    Great improvements over cyclegan! It's so much faster to train and occupies less gpu memory so I was able to train on larger images (than cyclegan). Thanks for thr detailed explanation, great work as always 👍

    • @connor-shorten
      @connor-shorten  4 года назад

      Thank you so much! That's really awesome to hear, thanks for sharing your experience!

  • @sayakpaul3152
    @sayakpaul3152 4 года назад +1

    Sweet! Special props for explaining patches.
    In the PacthNCE formulation, the addition of the representations from the other spatial locations has a great impact on the contrastive learning task here I think.

    • @connor-shorten
      @connor-shorten  4 года назад

      Thank you! Yeah, it reminds me of early style transfer algorithms with the gram matrix the way they compare features in intermediate layers.

    • @sayakpaul3152
      @sayakpaul3152 4 года назад +1

      Damn right!

  • @the-only-baka
    @the-only-baka 22 дня назад

    What about unpaired language translation? That would be cool!

  • @milanchatterjee1662
    @milanchatterjee1662 2 года назад

    Great video @Henry AI Labs Can you make a video on implementation of Contrastive Learning for Unpaired Image-to-Image Translation?

  • @mohamedabbashedjazi493
    @mohamedabbashedjazi493 4 года назад

    Let's consider an image of a synthetic hand, we want to translate it to a real hand, different patches in the source image are very similar, how do you eliminate the false positive patches in this case?

  • @NM-jq3sv
    @NM-jq3sv 4 года назад +1

    Won't making the embeddings of patches at similar locations same make the feature extractor learn style invariant or content features. Will this not make the discriminator bad ?

    • @connor-shorten
      @connor-shorten  4 года назад

      There are two separate losses here, the adversarial loss discriminator is separate from this. This loss is derived from a projection head after re-passing x and y' through the generator's encoder layers.

    • @NM-jq3sv
      @NM-jq3sv 4 года назад

      @@connor-shorten ah got it

  • @diegoantoniorosariopalomin4977
    @diegoantoniorosariopalomin4977 3 года назад +1

    great video