Tweetovi

Blokirali ste korisnika/cu @ikdeepl

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

  1. proslijedio/la je Tweet
    29. sij
    Poništi
  2. proslijedio/la je Tweet
    20. sij

    Given the smoothness of videos, can we learn models more efficiently than with ? We present Sideways - a step towards a high-throughput, approximate backprop that considers the one-way direction of time and pipelines forward and backward passes.

    Poništi
  3. proslijedio/la je Tweet

    This is hilarious. A Norwegian TV programme pitted some professional stock-pickers against an astrologist, some beauty bloggers and - i kid you not - cows shitting in a field. (short thread).

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

    We have released the PyTorch code of our method "HRNet + OCR + SegFix" at , which achieves the 1-st place on the leaderboard and competitive performance on 4 more benchmarks according to

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet

    How does deep learning perform DEEP learning? Microsoft and CMU researchers establish a principle called "backward feature correction" and explain how very deep neural networks can actually perform DEEP hierarchical learning efficiently:

    Poništi
  6. proslijedio/la je Tweet
    9. sij

    Nice piece in Quanta about gauge CNNs and geometric deep learning

    Poništi
  7. proslijedio/la je Tweet
    7. pro 2019.

    Datashader Release 0.9.0 is out, now supporting GPUs at last! Short story is that in our testing a NVIDIA GPU was 40X faster than single-core Pandas and 4.5x faster than multi-core Dask for rendering 100 million points. See for the details.

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

    A nice work on scale-equivariant CNNs via group convolutions. In contrast to e.g. SO(2), the dilation group (R^+,*) is non-compact. A conv over scales thus needs to be restricted to a certain range which introduces boundary effects similar to the zero padding artifacts of CNNs.

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

    Really pleased that our paper "Multiplicative Interactions and Where to Find Them" was accepted to . We present a unifying picture of several different neural net architectural motifs, that all involve multiplicative interactions... 1/3

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

    Decisions released 🎉 Congratulations to accepted papers; to those who we could not accommodate, we wish you success in your ongoing research. See our blog for the first of our reflections. See you soon in Ethiopia. 🇪🇹🌍

    Poništi
  11. proslijedio/la je Tweet
    17. pro 2019.

    Passionate about AI and wildlife conservation? Then you might be interested in the new Camera Trap Challenge from and AI for Earth. The top-performing solutions have a shot at prize money. Learn more:

    Poništi
  12. proslijedio/la je Tweet
    17. pro 2019.

    Episode EAI: AI Meets . In honour of this week's release, we put our OCR to work. OCR, or Optical Character Recognition, uses computer vision to recognize text. What can you do with our OCR?

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

    This will be the year remembered as when scientists captured the first-ever image of a black hole. It is among the choices for the best science photos of 2019.

    Poništi
  14. proslijedio/la je Tweet
    17. pro 2019.

    I am co-organizing MLSS-Indo 2020 in Indonesia. . Indonesia is a vast country with around 6,924,511 students but little papers in ML. I am looking for sponsors who want to support a student to attend the summerschool. Contact me if you are interested!

    Poništi
  15. proslijedio/la je Tweet

    “Australia's bushfires have emitted 250m tonnes of CO2, almost half of country's annual emissions.”

    Poništi
  16. proslijedio/la je Tweet
    14. pro 2019.

    Out: Winning a Turing Award for inventing the ConvNet In: Co-presenting invited talk @ via tweet. Lesson: “Lesson: don’t cancel your workshop talk last minute!” this morning w : Watch these posters get .

    Prikaži ovu nit
    Poništi
  17. proslijedio/la je Tweet
    14. pro 2019.

    Over 50 submissions to the climate change workshop at NeurIPS. Great! Keep going!

    Poništi
  18. proslijedio/la je Tweet
    12. pro 2019.
    Odgovor korisniku/ci

    Can I start a massive twitter war by positing that ideas from neuroscience have only done harm in machine learning? Neuro-terminology allows us to anthropomorphize technology that's more stupid than smart.

    Poništi
  19. proslijedio/la je Tweet
    11. pro 2019.

    Check out our poster #143 on general E(2)-Steerable CNNs tomorrow, Thu 10:45AM. Our work solves for the most general isometry-equivariant convolutional mappings and implements a wide range of related work in a unified framework. With

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

    Congrats to our winners of the Outstanding New Directions Paper Award at @NeurIPS2019 – & from CMU and . Find the download of the full paper “Uniform Convergence may be unable to explain generalization in Deep Learning” here:

    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:

    ·