Комментарии •

  • @daraburke
    @daraburke 6 месяцев назад

    Great overview of this new feature, thanks for taking the time to test and share the results.

  • @rodbotic
    @rodbotic 8 месяцев назад +1

    Thanks for bringing this to my attention. I am going to have to reprocess some of my previous models.

  • @DanFrederiksen
    @DanFrederiksen 8 месяцев назад +4

    Sharper but pixel resolution was never really the issue, we need smarter algos that capture the geometry better. Afaik there are no photogrammetry softwares that have moved into machine learning yet which is a pretty big revolution to overlook. A particularly grave example for these old gen algos is the trusses on the telescope that it just captures as blobs when a neural algo should recognize them as tubes with the exactitude our perception does. If you look at the antenna head it is atrociously bad. Imagine if that was crisp geometry like it was a cad model.

    • @OwenIverson
      @OwenIverson 5 месяцев назад

      "Afaik there are no photogrammetry softwares that have moved into machine learning yet" These are called NeRFs and they're incredible.

    • @DanFrederiksen
      @DanFrederiksen 5 месяцев назад +2

      @@OwenIverson well nerfs don't create 3D mesh yet so not quite there

    • @KeyFactory
      @KeyFactory 4 месяца назад

      @@OwenIversonis there already an NERF software for testing? I dont mean python repos, or stuff i need to compile first. thanks in advanced.

    • @OwenIverson
      @OwenIverson 4 месяца назад +1

      I honestly don't know - if I had to guess, I'd say it's probably still a lab-based thing (or if you're proficient in machine learning in general). It'll prob be a ways before we have a desktop app. @@KeyFactory

    • @KeyFactory
      @KeyFactory 4 месяца назад

      @@OwenIverson ok Thank you! I will wait :))