- 18
- 7 851
Yuxiang "Shawn" Wang
Приєднався 2 гру 2015
[UPDATED] ViViT & NaViT papers: How Sora encoded space-time patches | Shawn's ML Notes
Update on April 8th, 2023:
- Fixed missing narration on slide 25
- Added explanation for accuracy increase from upsampling (thanks to @ryuku4966!)
- Amplified audio track
Original video (archived): ua-cam.com/video/5-thXn708QA/v-deo.html
--
Thank you for checking out my video notes on ViViT & NaViT papers: how Sora encoded space-time patches! I would love to share my ML learning journey with you.
Paper information:
- Arnab, Anurag, et al. "Vivit: A video vision transformer." Proceedings of the IEEE/CVF international conference on computer vision. 2021.
- Dehghani, Mostafa, et al. "Patch n’pack: Navit, a vision transformer for any aspect ratio and resolution." Advances in Neural Information Processing Systems 36 (2024).
Please let me know in the comment section regarding any questions, points of discussion, or anything you would like see next. See you in the next video!
- Fixed missing narration on slide 25
- Added explanation for accuracy increase from upsampling (thanks to @ryuku4966!)
- Amplified audio track
Original video (archived): ua-cam.com/video/5-thXn708QA/v-deo.html
--
Thank you for checking out my video notes on ViViT & NaViT papers: how Sora encoded space-time patches! I would love to share my ML learning journey with you.
Paper information:
- Arnab, Anurag, et al. "Vivit: A video vision transformer." Proceedings of the IEEE/CVF international conference on computer vision. 2021.
- Dehghani, Mostafa, et al. "Patch n’pack: Navit, a vision transformer for any aspect ratio and resolution." Advances in Neural Information Processing Systems 36 (2024).
Please let me know in the comment section regarding any questions, points of discussion, or anything you would like see next. See you in the next video!
Переглядів: 430
Відео
Orignal transformer paper "Attention is all you need" introduced by a layman | Shawn's ML Notes
Переглядів 4,9 тис.Місяць тому
Thank you for checking out my video notes on the orignal transformer paper "Attention is all you need", as introduced by a layman - me! I would love to share my ML learning journey with you. Paper information: - Vaswani, Ashish, et al. "Attention is all you need." Advances in neural information processing systems 30 (2017). Please let me know in the comment section regarding any questions, poin...
![[ARCHIVED] ViViT & NaViT papers: How Sora encoded space-time patches | Shawn's ML Notes](http://i.ytimg.com/vi/5-thXn708QA/mqdefault.jpg)
[ARCHIVED] ViViT & NaViT papers: How Sora encoded space-time patches | Shawn's ML Notes
Переглядів 1,1 тис.Місяць тому
⚠️ An updated version of this video is available at: ua-cam.com/video/gpWCIgx3xMU/v-deo.html Update on April 8th, 2023 (only in the updated video, link above): - Fixed missing narration on slide 25 - Added explanation for accuracy increase from upsampling (thanks to @ryuku4966 !) - Amplified audio track Thank you for checking out my video notes on ViViT & NaViT papers: how Sora encoded space-ti...
Computational Mechanics Journal Club Session #6 - Generalized Eigenproblems Cont'd
Переглядів 883 роки тому
Link to the notes: drive.google.com/file/d/129yAilPdGGJ365K6EXmutcUc4NTCs6dZ/view?usp=sharing Welcome to the sixth session of our journal club on computational mechanics - generalized eigenproblems! In this session we will continue to go through some basic algorithms for generalized eigenproblems in structural analysis. Presenter: Dr. Daning Huang, Pennsylvania State University Links to related...
Computational Mechanics Journal Club Session #5 - Generalized Eigenproblems
Переглядів 693 роки тому
Topic: basic algorithms for generalized eigenproblems in structural analysis. Presenter: Dr. Daning Huang, Pennsylvania State University Links to related materials: en.wikipedia.org/wiki/Eigenvalues_and_eigenvectors en.wikipedia.org/wiki/Eigendecomposition_of_a_matrix#Generalized_eigenvalue_problem files.transtutors.com/cdn/uploadassignments/425722_1_203077-1-numerical-linear-aljebra.pdf (Part V)
Computational Mechanics Journal Club Session #4 Structural Dynamics
Переглядів 1353 роки тому
Welcome to the fourth session of our journal club on computational mechanics - structural dynamics! In this session we will touch upon the time stepping in structural dynamics problems. Presenter: Yuxiang "Shawn" Wang Links to lecture notes used: people.duke.edu/~hpgavin/cee541/NumericalIntegration.pdf
Computational Mechanics Journal Club Session #3 Beam Element
Переглядів 1694 роки тому
Paper: C0 Timoshenko Beam Element, MOOSE Documentation, mooseframework.inl.gov/modules/tensor_mechanics/C0TimoshenkoBeam.html. Bathe, K. J., & Bolourchi, S. (1979). Large displacement analysis of three‐dimensional beam structures. International journal for numerical methods in engineering, 14(7), 961-986. Speaker: Yuxiang "Shawn" Wang.
Computational Mechanics Journal Club Session #2 - Aerothermoelasticity
Переглядів 2154 роки тому
Paper: Huang, D., Friedmann, P. P., & Rokita, T. (2019). Aerothermoelastic Scaling Laws for Hypersonic Skin Panel Configurations with Arbitrary Flow Orientation. AIAA Journal, 1-16. Speaker: Daning Huang, Ph.D.
Computational Mechanics Journal Club Session #1 - Explicit Contact
Переглядів 5064 роки тому
Paper: Malone, J. G., & Johnson, N. L. (1994). A parallel finite element contact/impact algorithm for non‐linear explicit transient analysis: Part I-The search algorithm and contact mechanics. International journal for numerical methods in engineering, 37(4), 559-590. Speaker: Yuxiang "Shawn" Wang.
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 5/6
Переглядів 188 років тому
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 5/6
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 3/6
Переглядів 208 років тому
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 3/6
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 1/6
Переглядів 298 років тому
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 1/6
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 2/6
Переглядів 248 років тому
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 2/6
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 4/6
Переглядів 58 років тому
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 4/6
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 6/6
Переглядів 228 років тому
Blossom - UVa Chinese Art Performing Troupe 2011 Showcase 6/6
CMBBE 2015 Talk for the automatic algorithm in FEA for implementing hyperelastic materials
Переглядів 448 років тому
CMBBE 2015 Talk for the automatic algorithm in FEA for implementing hyperelastic materials
Three minute thesis preheat - Exploring the impact of skin mechanics on our sense of touch
Переглядів 368 років тому
Three minute thesis preheat - Exploring the impact of skin mechanics on our sense of touch
Hi I am very happy that I was able to find your channel on UA-cam I hope you will make more videos about computer vision keep Going ✌
Thank you for your support!
This is such a great explanation, do you plan to cover the "DiT: Scalable Diffusion Models with Transformers" paper sometime soon? Thanks a lot for such wonderful and insightful explanations...
Thank you for the kind words! That's a good idea and let me look into it. :)
Incredible content and your style is a perfect mix of confident and relatable. Keep it up!
I appreciate the encouragement :)
pretty okay until andrew's attention slide, then when it comes to your own explanations things become murky, and when you get "explain" the decoder, and then the full codec, you're swiping everything under the rug in a few short seconds when in fact this is exactly the section you should have spent most of time. all in all, a nice video until adrew's slide, basically worthless afterwards
Thanks for the feedback! Will learn to improve :) Would you mind explain in more details on which part I was missing for the encoder details? I can look into those and see if I can add some later!
@@yuxiangwang9624 darn, i got a notification that you responded to my comment, but only the first line of your reply was shown ("Thanks for the feedback! Will learn to improve :)"), and i didn't actually open to see your full reply until now. I will be back to you with the details, sorry for the delay...
one of the very few videos i found on youtube that explains the architecture very well
Thank you so much for the recognition!
Great voice. For fun, audition for a voice actor gig. Would look great on resume. Or on a date or at a conference. Lol
Lol thanks for the compliment!
Seems like a great video, subbed! 🙂
Thanks for the sub! Appreciate the recognition ❤️
please do more videos like this
Thank you! Will do :)
amazing video!
Thank you!
You aren't definitely a layman
Another Video! Looking forward to watching.
Haha thank you for your support! It was an old deck I made a year ago, so I might as well record it :)
amazing video.
Glad you liked it!
Is there some audio missing around 29:12? Nuancing the best positional embeding. Factorized(+)
Nice vid. Could be when you upscale it works better cause then its like the model is looking at smaller patches. An interesting ablation would have been to consider smaller patch size and check
Aha that's a good explanation! Makes perfect sense to me. I appreciate the reply & feel happy that I learned more through sharing!
Awesome video! I've always wanted to delve into ViT but haven't had the time. This video really did help reinforce my understanding, as well as add some really insightful details into all of these new methods. Thanks!
Thanks for appreciating and leaving a comment! :)
Loved this!
Thanks for your support!
有中文版么
之后可以录一个！
Thanks for sharing ♥
Thank you for your support! Please also feel free to leave a message and let me know the next topics you might be interested in. 😃
十年老粉，不请自来。不明觉历，催眠神器。
瑶山夜歌？？
我听她们说叫瑶族舞曲，应该是同一个东西？
给黄老师点赞！
赞
哈哈哈谢谢！！