Jia-Bin Huang

@jbhuang0604

Assistant professor at Virginia Tech, working on computer vision and machine learning

Blacksburg, VA
Vrijeme pridruživanja: svibanj 2015.

Tweetovi

Blokirali ste korisnika/cu @jbhuang0604

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @jbhuang0604

  1. proslijedio/la je Tweet
    prije 20 sati

    Some MIT faculty have put together a course called "The Missing Semester of Your CS Education." Having looked it over a bit, it looks fantastic and will benefit data science people from non-dev backgrounds fill in a lot of gaps, too.

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    28. sij

    The Spotlight this week is on Ph.D. student Jinwoo Choi who focuses on video understanding, a relatively under-explored but crucial research area. He will be sharing some of his work in March.

    Poništi
  3. 25. sij

    Code for our spotlight paper is available: Bonus: Our method also improves few-shot learners on standard single-domain few-shot classification tasks (with~ 2% improvement over MetaOptNet-SVM on mini-ImageNet).

    Poništi
  4. proslijedio/la je Tweet
    24. sij

    Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation by Hung-Yu Tseng et al. including

    Poništi
  5. 13. sij

    A weird feeling for advertising a CVPR19 paper at this point b/c it seems somehow outdated already (given the current pace of the field). BUT, we finally get the CrDoCo code ready! Enjoy!

    Poništi
  6. proslijedio/la je Tweet
    11. sij

    Know someone who wants to join twitter but isn’t quite sure how to get started? I was there not long ago and this column is for them. Thank you to and for the opportunity to contribute to !

    Poništi
  7. 1. sij

    I don't quite understand the role of journal editors. It seems to me that their goal is to mess up all the typesetting of the paper.

    Poništi
  8. 20. pro 2019.

    In ICLR last year, we show that existing few-shot classification methods perform poorly across different domains (). This year in , we show how we can make few-shot learners generalize better to unseen domains ()!

    Poništi
  9. proslijedio/la je Tweet
    18. pro 2019.

    Dive into the world of artificial intelligence in the season premiere of hosted by . Now streaming free on :

    Poništi
  10. 14. pro 2019.

    So glad that I attended ’s talk at today! In addition to the fascinating work on imagining the first picture of black hole, I am so impressed by the clarity of her talk!

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

    Ph.D. students Jinwoo Choi and attended with their advisor to present work on mitigating scene biases in action recognition.

    Poništi
  12. 12. pro 2019.

    Jinwoo and presenting our work on mitigating scene biases in action recognition.

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

    Why Can't I Dance in the Mall? Learning to Mitigate Scene Bias in Action Recognition by Jinwoo Choi et al. including

    Poništi
  14. 11. pro 2019.

    The arXiv link here:

    Prikaži ovu nit
    Poništi
  15. 10. pro 2019.

    Please come to our poster on Thursday night! Jinwoo, , and Joseph will tell you how can we mitigate the spurious correlation between the scene and action. Project/code:

    Prikaži ovu nit
    Poništi
  16. 6. pro 2019.
    Poništi
  17. 2. pro 2019.

    DRIT++ accepted to IJCV! While journal pubs take longer than conferences, I find it rewarding to go through the revision processes to polish the work. Reviewers - in journals: How can I help improve this work? - in confs: How can I kill this paper?

    Poništi
  18. proslijedio/la je Tweet
    1. pro 2019.

    Oh, also: you know how your 8 year old self was mesmerized with the night sky using your backyard telescope? Applying to grad school is not about that. And yet, to much a committee's dismay, 40-50% of personal statements start that way. Just ...don't.

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

    HOW TO GIVE A CONFERENCE TALK IN FIVE EASY STEPS 1. Write the outline! It'll help you collect your thoughts and make a good talk!!! 2. Write a first draft! Doesn't need to be that good, just get your thoughts down!!! 3. Rehearse! Read the draft aloud and see if it sounds good!

    Prikaži ovu nit
    Poništi
  20. proslijedio/la je Tweet
    24. stu 2019.

    I've learned a few big lessons * Making a thing practical is harder than you think * You'll hear <it'll never work> 1000 times. It's OK * Don’t dismiss good ideas because they’re not yet practical * There’s never a “main” inventor - it’s always a team sport * It pays to be lucky

    Prikaži ovu nit
    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:

    ·