Mohamed Hamed

@hamed_mo7amed

Ph.D. candidate working on Computer Vision and Deep learning. I mostly tweet about machine learning and deep learning.

Vrijeme pridruživanja: studeni 2017.
13 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @hamed_mo7amed

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @hamed_mo7amed

  1. Prikvačeni tweet
    1. velj

    As a researcher what is the best IDE you use for Python when you use Pytorch to debug your network. Feel free comment down below with the one not listed.

    2 korisnika označavaju da im se sviđa
    Poništi
  2. proslijedio/la je Tweet
    4. velj

    A research story with a twist I looked for a neural network regularization method that limited the number of non-zero node activities (L0 group sparsity?) I couldn’t find one. 1/11

    6 replies 32 proslijeđena tweeta 168 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet
    4. velj

    I didn't discover this feature until today. This should be the default at every presentation!

    7 replies 123 proslijeđena tweeta 585 korisnika označava da im se sviđa
    Poništi
  4. 4. velj

    ... and yes Education is irrelevant. No matter if you’re doing master or PhD you have to pass the hardcore coding test. hmmmmm

    Our NN is initially in Python for rapid iteration, then converted to C++/C/raw metal driver code for speed (important!). Also, tons of C++/C engineers needed for vehicle control & entire rest of car. Educational background is irrelevant, but all must pass hardcore coding test.
    Prikaži ovu nit
    1 korisnik označava da mu se sviđa
    Poništi
  5. proslijedio/la je Tweet
    3. velj

    - Unofficial PyTorch implementation of GLAMpoints: Greedily Learned Accurate Match points - ported weights + re-trained model on PS-Dataset.

    12 proslijeđenih tweetova 32 korisnika označavaju da im se sviđa
    Poništi
  6. proslijedio/la je Tweet
    1. velj

    Because I’m starting to turn to twitter for advice (rather than just moaning), suggestions on how to organise/store ideas? Having lots of “could do that” ideas atm- how do others manage to not forget, but not deal with instantly (because I’m collecting postits😂)

    46 replies 9 proslijeđenih tweetova 73 korisnika označavaju da im se sviđa
    Poništi
  7. proslijedio/la je Tweet
    31. sij

    On Tuesday, in my class, we have learnt that all a neural net does is stretching / contracting the space fabric. For example this 3-layer net (1 hidden layer of 100 positive neurons) gets its 5D logits (2D projections) linearly separable by the classifier hyperplanes (lines).

    202 proslijeđena tweeta 1.058 korisnika označava da im se sviđa
    Poništi
  8. proslijedio/la je Tweet
    31. sij

    An Opinionated Guide to ML Research: “To make breakthroughs with idea-driven research, you need to develop an exceptionally deep understanding of your subject, and a perspective that diverges from the rest of the community—some can do it, but it’s hard.”

    9 replies 238 proslijeđenih tweetova 858 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  9. 31. sij

    Thanks @facbookai for making our life as a researcher much easier.

    0:15
    We are releasing HiPlot, a lightweight interactive visualization tool to help AI researchers discover correlations and patterns in high-dimensional data.
    1 proslijeđeni tweet 6 korisnika označava da im se sviđa
    Poništi
  10. proslijedio/la je Tweet
    29. sij

    New blog post: Contrastive Self-Supervised Learning. Contrastive methods learn representations by encoding what makes two things similar or different. I find them very promising and go over some recent works such as DIM, CPC, AMDIM, CMC, MoCo etc.

    8 replies 99 proslijeđenih tweetova 452 korisnika označavaju da im se sviđa
    Poništi
  11. proslijedio/la je Tweet
    29. sij

    PolyGame: An open-source framework for training AI players through self-play. Deals with many games, board size variation, partial observability... Interesting generalization tidbit: It plays Go on 19x19 at very good level after training only on 13x13.

    5 replies 122 proslijeđena tweeta 349 korisnika označava da im se sviđa
    Poništi
  12. 28. sij

    Nested for loop for tensors lol

    0:05
    cubes within cubes within cubes within... (©:)
    1 reply 2 korisnika označavaju da im se sviđa
    Poništi
  13. proslijedio/la je Tweet
    27. sij

    Kornia v0.2.0 is out ! We have introduced a new data augmentation module with strong GPU support, extended the set of color conversion algorithms, supporting GPU CI tests with v1.4.0, and much more. Happy coding !

    162 proslijeđena tweeta 518 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  14. 27. sij

    I like the question and like the reply more!

    "Maximum likelihood learning of the data generating distribution via ELBO is biased if p(z|x) is not in the variational family." What is the right reference?
    1 korisnik označava da mu se sviđa
    Poništi
  15. proslijedio/la je Tweet
    26. sij

    Teaching Deep Unsupervised Learning (2nd edition) at this semester. You can follow along here: Instructor Team: , , , Wilson Yan, Alex Li, YouTube, PDF, and Google Slides for ease of re-use

    Korisnici , i
    121 proslijeđeni tweet 413 korisnika označava da im se sviđa
    Poništi
  16. proslijedio/la je Tweet
    24. sij

    I can recommend this book (~200 p.), which I red during my PhD time, and which fixes the biggest knowledge gaps:

    8 proslijeđenih tweetova 54 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  17. proslijedio/la je Tweet
    24. sij

    is one of the most important techniques I don't often recommend PhD Thesis' - 's is exceptional. He's a brilliant writer! Check out this taxonomy / table of contents!!! 👇👇👇

    7 replies 297 proslijeđenih tweetova 1.033 korisnika označavaju da im se sviđa
    Poništi
  18. proslijedio/la je Tweet
    24. sij

    Facebook's AI Residency program, which gives residents 12 months of practical experience in AI working with leading researchers, is accepting applications until January 31st! Current residents Diana and Eric discuss their experience in this Q&A:

    8 replies 47 proslijeđenih tweetova 159 korisnika označava da im se sviđa
    Poništi
  19. proslijedio/la je Tweet
    24. sij

    When I invented adversarial training as a defense against adversarial examples, I focused on making it as cheap and scalable as possible. Eric and collaborators have now upgraded the original cheap version to compete with newer, more expensive versions.

    1/ New paper on an old topic: turns out, FGSM works as well as PGD for adversarial training!* *Just avoid catastrophic overfitting, as seen in picture Paper: Code: Joint work with and to be at
    Prikaži ovu nit
    8 replies 120 proslijeđenih tweetova 679 korisnika označava da im se sviđa
    Poništi
  20. 24. sij

    Training without Backprop and labeled data, that’s interesting!

    Backpropagation and labeled data are the bread and butter of deep learning. But recent research from the University of Amsterdam suggests neither is necessary to train effective neural networks to represent complex data:
    1 proslijeđeni tweet 3 korisnika označavaju da im se sviđa
    Poništi

@hamed_mo7amed još nije objavio nijedan Tweet.

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:

    ·