David Güera

@dgueraco

 Research Scientist | AI/ML. Computer vision, AR/VR, and machine learning.

Seattle, WA
Vrijeme pridruživanja: srpanj 2009.

Tweetovi

Blokirali ste korisnika/cu @dgueraco

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @dgueraco

  1. proslijedio/la je Tweet
    4. velj

    The names of 100 prisoners are put in 100 boxes. One at a time, privately, each prisoner opens and closes 50 boxes. They are set free if *everyone* finds their own name. Strategy can be discussed beforehand, but otherwise there's no communication. What's the best strategy? 1/23

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    2. velj

    I actually found this behind the scenes of Porsche’s Super Bowl Taycan commercial more interesting than the commercial or the Super Bowl itself. Enjoy, fellow video nerds 🤓

    Poništi
  3. proslijedio/la je Tweet
    2. velj
    Poništi
  4. proslijedio/la je Tweet
    1. velj

    On a train in Italy. A teenage Chinese boy boards the train. A woman comments loudly: “There you go, we are all going to be infected.” He replies in perfect Italian with a Roman inflection: “Ma’am, in my whole life I’ve seen China only on google maps.” Applauses.

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    1. velj

    This is the crankshaft for a Wärtsilä-Sulzer RTA96-C engine, the largest reciprocating engine in the world, used in large container ships. It's a 1810-liter engine that generates 108,920 horsepower at 102 RPM. This crankshaft weighs 300 tons

    Poništi
  6. proslijedio/la je Tweet

    Es confirma que cap de les pel·lícules d'Studio Ghibli que ha pujat avui Netflix inclou el doblatge en català malgrat que el doblatge en català ja està fet en moltes d'aquestes pel·lícules.

    Poništi
  7. proslijedio/la je Tweet
    30. sij

    Announcing the 3rd Workshop & Challenge on Learned Image Compression at , which challenges researchers to use techniques to improve the quality and lower the bandwidth needed for multimedia transmission. Learn all about it ↓

    Poništi
  8. proslijedio/la je Tweet

    Wrote this simple app to break up larger decisions into smaller pairwise comparisons (using Trueskill to draw out preferences): Wrote it to find a name for the baby we're expecting in a few weeks and found it really helpful. Hope others do as well.

    Prikaži ovu nit
    Poništi
  9. proslijedio/la je Tweet
    25. sij

    The reason Google’s new search design can claim to “boldy” label ads while making them difficult to distinguish has to do with something called pre-attentiveness.

    Prikaži ovu nit
    Poništi
  10. proslijedio/la je Tweet

    The is set for its first flight this morning. Follow the flight live shortly and learn what makes the 777X different at

    Prikaži ovu nit
    Poništi
  11. proslijedio/la je Tweet

    LIVE 👏 CODING 👏 INTERVIEWS 👏 ARE 👏 WORTHLESS 👏👏👏👏👏👏👏

    Poništi
  12. proslijedio/la je Tweet
    22. sij

    It took us 255 years to go from Galilean transformations for falling bodies to Lorentz transformations that are in accordance with Einstein's special relativity.

    Poništi
  13. proslijedio/la je Tweet
    20. sij

    El pobre iaio de 92 anys d'avui. "Segons han explicat tant un familiar de l’afectat com la propietària del pis, el Cinto té diversos pisos de propietat i recentment s’ha venut una finca per 400.000 euros."

    Prikaži ovu nit
    Poništi
  14. proslijedio/la je Tweet
    21. sij

    Many thanks to a tip from , I've added a brief section to my self-supervised learning post on noise contrastive estimation and consistency loss. And thanks to I've also added a paragraph about the history of self-supervised learning

    Prikaži ovu nit
    Poništi
  15. proslijedio/la je Tweet
    19. sij

    ⏲️ As of today, we have about eighteen years to go until the Y2038 problem occurs. But the Y2038 problem will be giving us headaches long, long before 2038 arrives. I'd like to tell you a story about this.

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    20. sij

    A pretty comprehensive review of detection methods. Definitely a lot more to be done but every great research starts with a good understanding of the state-of-the-art. Read, enjoy and share!

    Poništi
  17. proslijedio/la je Tweet
    20. sij

    1/ This time I just have a single image here. It was taken in 1900 in New York City, and it's of course colorized by my latest and greatest and unreleased DeOldify model. I've done this one before but it's great to test the model on high resolution renders.

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet
    18. sij

    A thread of classifiers learning a decision rule. Dashed line is optimal boundary. Animations with by and . Logistic regression {stats::glm} with each class having normally distributed features. (1/n)

    Prikaži ovu nit
    Poništi
  19. proslijedio/la je Tweet
    16. sij

    Hi, I'm Jess 👋 I've been let go in the Mozilla layoffs. I'm now free 2 days/week for contract work in developer relations, management support, hiring or strategy for startups. FAQ in thread. RTs or leads so welcome💖

    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·