Andrei Bursuc

@abursuc

Computer Vision and Machine Learning research scientist at Safran Tech; Inria alumni; PhD from Mines ParisTech

Paris, France
di.ens.fr/~bursuc
Joined November 2008
167 Photos and videos Photos and videos

Tweets

You blocked @abursuc

Are you sure you want to view these Tweets? Viewing Tweets won't unblock @abursuc

  1. Jun 9
    Replying to

    With 352 P100 GPUs, this implementation on ResNet-50 does one pass of all 1.28M ImageNet images in ~30 seconds

    10 likes
  2. Jun 8

    Low-shot learning with large-scale diffusion When et al. use "large-scale", they really mean it: 100M images

    6 retweets 19 likes
  3. Jun 8

    Currently, there are no downsides for flag-planting, only potential benefits. They should involve risk/responsibility to face such critics.

    So I took some time and wrote this medium post about this language generation paper.
    1 retweet 2 likes
  4. Jun 8

    So in the end Softbank will be buying Boston Dynamics. Looking forward to see the Japanese touch in their robots

    1 reply 4 likes
  5. Jun 8

    Training with mini-batches of 8192 images. Wowowow!

    Training ImageNet in 1 hour with 256 GPUs, minibatches of 8192. From Facebook's FAIR: 1 hour... incredible
    2 replies 4 retweets 16 likes
  6. Jun 6

    See, Hear, and Read: what happens when you align vision, sound and text modalities Nice work by et al.

    4 retweets 12 likes
  7. Jun 5

    And the magsafe ...

    All I wanted from was my escape key back.
    1 like
  8. Retweeted
    Jun 3
    Replying to and

    Actual footage of my advisor vs me trying to get a job

    10 replies 615 retweets 1,528 likes
  9. Retweeted
    Jun 3

    The CfP for our Video Games and Machine Learning workshop at ICML2017 has arrived (cc )

    46 retweets 82 likes
  10. May 31

    Inside-out tutorial and visualization of LSTMs

    7 retweets 20 likes
  11. Retweeted
    Apr 6

    I think it’s time to coin the term “Artificial Artificial Intelligence” for products that claim to use ML but really don’t.

    3 replies 42 retweets 64 likes
  12. May 30
    Replying to

    Check out also this related work Unsupervised monocular depth estimation with left-right consistency

    1 reply 4 retweets 6 likes
  13. May 30
    Replying to

    ... brought to you by Zhu, Brown, Snavely and Lowe. Results are quite impressive:

    1 reply 3 retweets 1 like
  14. May 30

    Unsupervised/self-supervised learning from video sequences for monocular depth and camera motion estimation

    1 reply 12 retweets 35 likes
  15. May 30

    A tour of PyTorch internals

    3 retweets 7 likes
  16. Retweeted
    Apr 21

    Multilingual word vectors in 78 languages

    2 replies 34 retweets 65 likes
  17. Retweeted
    May 25

    Hilarious new post from "Everything that Works Works Because it's Bayesian: Why Deep Nets Generalize?"

    8 replies 128 retweets 243 likes
  18. Retweeted
    May 23

    My first paper at DeepMind: What can be learnt by looking at and listening to a large amount of unlabelled videos?

    6 replies 148 retweets 312 likes
  19. May 23

    Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset + a zoo of proposed and evaluated architectures

    4 retweets 12 likes
  20. May 22
    Replying to

    This is a useful reminder for reviewers from the program chairs

    2 likes

@abursuc hasn't Tweeted yet.

Loading seems to be taking a while.

Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

    You may also like

    ·