Ming-Yu Liu

@liu_mingyu

Working towards enabling machine super-human imagination capability. Distinguished Research Scientist at NVIDIA. Tweets are my own.

San Jose, CA
Vrijeme pridruživanja: prosinac 2015.

Tweetovi

Blokirali ste korisnika/cu @liu_mingyu

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

  1. Prikvačeni tweet
    13. lip 2019.

    The beta version is now available to everyone as a web service via AI Playground A short illustration video is available in May everybody have fun with the app. ,

    Poništi
  2. proslijedio/la je Tweet
    1. velj
    Odgovor korisnicima

    I just went through ~3 months of pain evaluating applications to the faculty search. Beyond some lower bound, pub count had almost no correlation with how we ranked the candidates. For papers, we mainly looked at the 3 representative publications included in the package.

    Poništi
  3. proslijedio/la je Tweet
    17. sij

    The AI for Content Creation workshop at CVPR 2020 is now accepting submissions! Many great guest speakers! Organized with Weilong Yang () and Kalyan Sunkavalli ().

    Poništi
  4. proslijedio/la je Tweet
    16. sij

    (1) Dear friends! Please help me spread the word! Proud to announce my first single author paper, a new display technology that combines optical scanning and machine learning. The results is up to x4 resolution, faster refresh rate and no more screen door effect!

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    14. sij
    Odgovor korisnicima

    I believe the term 'self-supervised learning' is actually quite old. By the 1990s it was pretty common (e.g., , ), and the earliest usage I see on Google Scholar is this 1978 paper:

    Poništi
  6. proslijedio/la je Tweet
    7. sij

    Tacotron2+WaveGlow TTS deployed through ONNX and TensorRT! Research at NVIDIA matters because we get to work as one big, company-wide team to translate ideas into impact. And then we get to participate in the broader community by releasing code, models, examples - like this:

    Poništi
  7. proslijedio/la je Tweet
    6. sij

    I'm proud of our Esports research on high framerate, low latency that contributed to the new NVIDIA G-SYNC Esports Display launch. Playing at 360 Hz feels simply amazing and makes a measurable difference in performance.

    Poništi
  8. proslijedio/la je Tweet
    21. pro 2019.

    I don't think there'll be another AI winter, because search, planning and supervised learning work well and are in use everywhere. There might be an RL winter though, because reinforcement learning doesn't work well and is not likely to work well soon.

    Prikaži ovu nit
    Poništi
  9. proslijedio/la je Tweet
    18. pro 2019.

    very excited to share UNAS, my first project . UNAS unifies differentiable and RL-based NAS, enabling us to search with both diff. & non-diff. loss. UNAS also proposes a generalization based loss to discover networks prone to overfitting. paper:

    Prikaži ovu nit
    Poništi
  10. 11. pro 2019.
    Poništi
  11. proslijedio/la je Tweet
    10. pro 2019.

    想学有所用,掌握 技能、落地 AI 场景应用?NVIDIA 深度学习学院()手把手教您如何用深度学习方法,实现自动化工业产品缺陷检测。12 月 17 日 苏州 China,《深度学习—— 工业检测》实战培训,名额有限,报名从速!

    Poništi
  12. proslijedio/la je Tweet

    Meet the AI choreographer! To help automatically create a dance video, researchers developed a model that can automatically compose new dance moves. The work is being presented at this week.

    Poništi
  13. proslijedio/la je Tweet
    4. pro 2019.

    Glad to share our new work, . Our model can generate high-quality images reflecting the diverse styles (e.g., hairstyles, makeup) of reference images. arXiv: github: co-authors:

    Prikaži ovu nit
    Poništi
  14. 3. pro 2019.

    Woohoo! GauGAN won the Best of What's New Award by Popular Science Magazine!!! If you haven't tried GauGAN, please visit

    Poništi
  15. proslijedio/la je Tweet
    13. stu 2019.

    Excited to release a library for 3D deep learning! Check it out, and give us feedback! Great effort by Edward Smith, JF Lafleche Artem Rozantsev, Tommy Xiang, Gav State, . We plan to extend it with many exciting features!

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    6. stu 2019.

    Dancing to Music This paper looks at generating dance from music through a decomposition to composition learning framework. Paper Project

    Poništi
  17. proslijedio/la je Tweet
    3. stu 2019.

    When people ask me what is the future of painting or brushes, I tell This is. This shows the power of and working together to enhance human creativity. I extracted and processed the video using the SPADE algorithm and augmented it on the feed.

    Prikaži ovu nit
    Poništi
  18. 1. stu 2019.

    Dusan’s talk on the fundamentals and NVIDIA AMP library for easy mixed precision training.

    Prikaži ovu nit
    Poništi
  19. proslijedio/la je Tweet
    1. stu 2019.

    Work we did on visualizing memorability, using GANs!

    Poništi
  20. 1. stu 2019.

    We will being giving a tutorial on how to use mixed precision (fp32 + fp16) for training deep networks for vision applications at in the Saturday morning. . Want to train your network faster with your current GPUs? Please come to check it out.

    Prikaži ovu nit
    Poništi
  21. 31. lis 2019.

    We will be doing a poster session for our paper on FUNIT: Few-shot Unsupervised Image-to-Image Translation () at Hall 1B. Poster 139. Hope to see you there.

    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:

    ·