RunwayML

@runwayml

Machine learning for creators. We're hiring:

New York, USA
Vrijeme pridruživanja: svibanj 2018.

Tweetovi

Blokirali ste korisnika/cu @runwayml

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

  1. Prikvačeni tweet
    4. lip 2019.

    Hello world, we have some exciting news to share with you today! 🔥🔥 Runway is now in public beta, available for anyone to download! 🔥🔥 Along with that, we are launching a brand new identity and website, check them out! ✒️ 💅 Learn more at

    Poništi
  2. proslijedio/la je Tweet
    prije 5 sati
    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet
    prije 19 sati
    Poništi
  4. 30. sij

    Learn how to train your own models in RunwayML! 💫 In this quick video tutorial, demonstrates how you can train your own generative image model in just a few clicks. 🎥

    Poništi
  5. proslijedio/la je Tweet
    27. sij

    Browsing through images in latent space in has a very multi-worlds / Amber Saga pattern walking feel to it -- a very fun user experience!

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

    Used to to extract depth frames from Koyaanasquatsi and re-project (naively) via SceneKit:

    Prikaži ovu nit
    Poništi
  7. proslijedio/la je Tweet

    Recorded this demo of installing and using for my remote Artificial Images class. With all the new models available in it these days I discuss how I would go about searching for new models.

    Poništi
  8. proslijedio/la je Tweet
    23. sij

    has a really easy way to train StyleGAN models now -- took me 5 mins to input 5k generated small molecule images and then hit the train button. Definitely beats the colab notebooks. These scribbles sort of look like someone trying to solve an NMR structure 😁

    Prikaži ovu nit
    Poništi
  9. 21. sij

    We are excited to announce that Model Training is now available to all RunwayML users! 🚂💫 Give it a try! 👉 We can't wait to see what you all create. Of course, we're just getting started. 🚀

    Poništi
  10. proslijedio/la je Tweet
    19. sij

    Like haunted by Kafka. An animation of the learned latent space from my first StyleGAN experiment adapting a model trained to generate beetles to 1235 kanji. Next: explore further via colour tagging ala

    Poništi
  11. proslijedio/la je Tweet
    19. sij

    本日のData Driven Artの授業は によるメディアアートの話と後半は2bitによる のハンズオンです。

    Poništi
  12. proslijedio/la je Tweet
    18. sij

    Matcap Generator made with . A weird little idea I had that seems to work pretty well.

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

    Also, training via is so easy! No setting up of infra, servers/EC2/Colab. Great at quicker feedback loops and getting something done fast.

    Prikaži ovu nit
    Poništi
  14. 16. sij

    Our video tutorial series on Generative Models continues! This week, talks about 's GLOW: a reversible generative model Check it out: 👀👇

    Poništi
  15. proslijedio/la je Tweet
    15. sij

    Few-shot face translation video, made with . 🔫💥

    Poništi
  16. proslijedio/la je Tweet

    One of my favorite ways to use is to try out new potentially SOTA models before spending an hour setting them up on a server. (Also shout out to Sree Harsha who seems to be the one porting all of them). Porting SRFBN to a Colab file for superres work now.

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

    Just for fun, I did an experiment where I shot my own training material. These images were generated from a library of 700 feathers, twigs & fluffy catkins photographs I took and fed to , which is now spitting them out in thousands of new arrangements.

    Poništi
  18. proslijedio/la je Tweet
    14. sij

    We need an entire industry building tools to give journalists like this superpowers. This example gives me so much hope for ; you could automate this in minutes. Actually migrated cnOCR to the platform if you're looking for Chinese character OCR

    Poništi
  19. proslijedio/la je Tweet
    13. sij

    Last week we gave a workshop OPENRNDR x Machine Learning at the Tech Week . In just four days the students got to know not only the basics of OPENRNDR but also experimented with some of the exciting possibilities and make possible.

    Prikaži ovu nit
    Poništi
  20. proslijedio/la je Tweet
    10. sij
    Odgovor korisnicima

    So is an app (for computers) that provides a user-friendly way to interact with 'AI' without code. One of the modules it makes available is DeOldify, a neural network for colouring black and white photos. Very straightforward and I ran through hundreds in minutes.

    Poništi
  21. proslijedio/la je Tweet
    8. sij

    Sneak peek at a styleGAN experiment using a data set of bacterial growth w/ made with the training GUI from

    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:

    ·