Rezultati pretraživanja
  1. 18. lip 2019.

    Released at , MediaPipe is Google's new framework for media processing pipelines, combining model-based inference via TensorFlow with traditional CV tasks like optical flow, pose tracking, and more. Used in existing projects like Motion Stills.

    7 replies 67 proslijeđenih tweetova 249 korisnika označava da im se sviđa
  2. 19. lip 2019.

    Now available for commercial use on NGC, pix2pixHD developed by researchers generates high-resolution photorealistic images from high-level labels.

    90 proslijeđenih tweetova 246 korisnika označava da im se sviđa
  3. 10. svi 2019.

    When video description meets bounding boxes! We are now releasing data/code/models/leaderboard (yes, everything) on our oral paper Grounded Video Description: Paper: Dataset (158k bboxes on 52k captions):

    1 reply 11 proslijeđenih tweetova 33 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
  4. 19. lip 2019.

    If you're at kindly visit me present our work on "Cycle Consistency for robust VQA" in Grand Ballroom at 14.10. My co-authors will be there to answer all your questions in the poster session #184 that follows. Dataset + Code + Paper:

    8 proslijeđenih tweetova 30 korisnika označava da im se sviđa
  5. 20. lip 2019.

    Ending my conference as a Crazy Rich Bayesian! Go Bayesians!

    28 replies 195 proslijeđenih tweetova 2.131 korisnik označava da mu se sviđa
  6. 17. lip 2019.

    If you enjoy 🍻 AND realtime visual-inertial SLAM, is hosting a happy hour tomorrow at 5pm. DM me for details!

    1 reply 12 proslijeđenih tweetova 43 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
  7. 20. lip 2019.

    I liked the DeepMapping paper from Ding and Feng at . A bit similar to DIP, they use deep learning machinery to solve a surprising optimisation problem (no learning on a dataset): pose graph alignment for a set of pose scans.

    40 proslijeđenih tweetova 126 korisnika označava da im se sviđa
  8. 18. lip 2019.

    Best paper award at main idea: seeing around the corner at non-line-of-sight (NLOS) objects by using Fermat paths, which is a new theory of how NLOS photons follow specific geometric paths.

    1 reply 34 proslijeđena tweeta 126 korisnika označava da im se sviđa
  9. 19. lip 2019.

    We have a fantastic representation at this week! 🌴🌊🌞 Come and say 'HI' to Marco, Lewis, Jamie, Simon, Miroslaw, Josef, John, Armin, Adrian, Tao, and Yi-Zhe! 👋👋👋

    4 proslijeđena tweeta 19 korisnika označava da im se sviđa
  10. 20. lip 2019.

    Gucci has nothing on this $13,000 hat

    7 proslijeđenih tweetova 38 korisnika označava da im se sviđa
  11. 19. lip 2019.

    If you have any doubts about the applicability of CNNs to local feature detection for robust pose estimation, come to our Deep Charuco poster today at . This videos shows a shadow effect and traditional methods are simply too frail. Red=fail

    27 proslijeđenih tweetova 87 korisnika označava da im se sviđa
  12. 17. srp 2019.

    Deep Flow-Guided Video Inpainting By Completing a missing flow is easier than filling in pixels of a missing region directly. SoTA on DAVIS and YouTube-VOS Code ArXiv

    28 proslijeđenih tweetova 69 korisnika označava da im se sviđa
  13. 15. lip 2019.

    Learn about Learning from Unlabeled Videos at , Sunday in Room E, 9:00am Fresh posters and keynotes: Antonio Torralba, Noah Snavely, Andrew Zisserman, Bill Freeman, Abhinav Gupta, Kristen Grauman

    1 reply 10 proslijeđenih tweetova 59 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
  14. 15. lip 2019.

    We willl be presenting our work "Divide and Conquer the Embedding Space for Metric Learning" at on Tuesday 18th: Poster 24 at 10:15. Authors: Me, Vadim Tschernezki, Uta Büchler () and Björn Ommer. Paper and Code:

    28 proslijeđenih tweetova 73 korisnika označavaju da im se sviđa
  15. 19. lip 2019.

    Y. Niitani of Preferred Networks gave an oral presentation at "Sampling Techniques for Large-Scale Object Detection from Sparsely Annotated Objects" used to win 2nd prize @ Google AI Open Images challenge (w/ Akiba, Kerola, Ogawa, Sano & Suzuki)

    10 proslijeđenih tweetova 39 korisnika označava da im se sviđa
  16. 18. ruj 2019.

    For those interested in some of the latest self-driving research (especially those participating in our competition), we’ve collected our favorite self-driving content from in our latest reader's digest. Read them here:

    35 replies 525 proslijeđenih tweetova 4.655 korisnika označava da im se sviđa
  17. 20. lip 2019.

    Pleased to now also share the code for our paper: "Meta-learning Convolutional Neural Architectures for Multi-Target Concrete Defect Classification with the CODEBRIM Dataset". Code: Paper: Dataset:

    10 proslijeđenih tweetova 21 korisnik označava da mu se sviđa
  18. 25. lip 2019.

    All papers with code at

    14 proslijeđenih tweetova 19 korisnika označava da im se sviđa
  19. 24. srp 2019.

    We're launching a blog for sharing our latest work and reflections on research for safer and increasingly autonomous vehicles. The first post describes our recent paper ADVENT for unsupervised domain adaptation:

    10 proslijeđenih tweetova 28 korisnika označava da im se sviđa
  20. 15. lip 2019.

    NVIDIA Research will present 20 papers at , including 11 orals. Full list is here: . Check it out!

    13 proslijeđenih tweetova 77 korisnika označava da im se sviđa

Nema rezultata.

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.