Deep Learning: A Crash Course (2018) | SIGGRAPH Courses
ฝัง
- เผยแพร่เมื่อ 13 พ.ค. 2024
- Deep learning is a revolutionary technique for discovering patterns from data. We'll see how this technology works and what it offers us for computer graphics. Attendees learn how to use these tools to power their own creative and practical investigations and applications.
Originally presented at SIGGRAPH 2018.
PRESENTER
Andrew Glassner, The Imaginary Institute
LEARN MORE
ACM DIGITAL LIBRARY: dl.acm.org/doi/10.1145/330536...
ACM SIGGRAPH: www.siggraph.org/ - ภาพยนตร์และแอนิเมชัน
![[1hr Talk] Intro to Large Language Models](http://i.ytimg.com/vi/zjkBMFhNj_g/mqdefault.jpg)
![BOWKYLION - ส่วนต่าง (do it without me) [Official MV]](http://i.ytimg.com/vi/7x-pwcMopzE/mqdefault.jpg)
i fall asleep with youtube on and now im waking up to this lmao
lmaoo same
Swear to God same
Same lol
Indeed! Me too
me too
In 2024, Don't set new year financial goals without consulting a financial adviser. Their expertise ensures a solid plan for success. Building wealth involves developing good habits like regularly putting money away in intervals for solid investments.
Thanks for the advice! I'm new to financial planning and wasn't sure where to start. Any tips on finding a reliable financial adviser or resources to guide beginners?
I completely understand the impact of the new year on financial decisions-it's transformative. While I've found success in investing, especially with the guidance of my advisor, Lisa Angelique Abel, I still appreciate the solid foundation i had. This principles align with the core values of financial discipline.
How can I get in touch with your financial consultant?
Just take a look at her full name on the internet. She is well known so it shouldn't be hard to find her.
I looked up her full name online and found her page. I emailed and made an appointment to talk with her; hopefully, she gets back to me.
If I had to choose only 1 video to watch on Deep Learning on TH-cam I would, without a doubt, pick this video. Superb teaching style.
i watched food Theory and sleept away know im 3hours inside deep learning
Beautiful
I literally had the same thing happen to me 😂
Same thing
YOO SAME BUT I SHARED THIS VID TO LIKE 30 RANDOM PEOPLE
im 1:47:57 learning very deeply this morning
Thank you Professor, for saving the time to browse the topics individually online
Felt like a 3.5 hour Ted Talk... that guy is such a rare, damned good, CS teacher... wow.
I opened the video by mistake and now I can not leave. Good presentation!
Imagine if all profs were these enthusiastic...
Been watching, listening, and paying attention to all of predictions and forecasts since early Covid. He hasn't disappointed yet 👌
Well said! I am also here to learn how to invest after listening to a lady on tv talk about the importance of investing and how she made 7 figures in 3 months, somehow the video taught me nothing and left me even more confused, I'm a newbie and I'm open to ideas on how to invest for retirement
@@BrunoLuke
lookup KRISTIN GAIL CUNNINGHAM , this is her name online, she's the real investment prodigy since the crash and has helped me recover my loses
@@MarcusFred-wn3iv Despite the economic crisis and the rate of unemployment now is the best time to invest
@@BrunoLuke Investment now will be wise but the truth is investing on your own will be high risk. I think it will be best to get a professional👌
@@MarcusFred-wn3iv Thank you, Going through her profile on her webpage out of curiosity, and surprisingly she seems proficient. I appreciate this.
single best thing I‘ve seen on the internet today (after watching 3 hours there wasn‘t much time left to browse the others).
one day will be able to swim and travel in deep learning sea
sarcasm made me press back
@@macdeep8523 oooooooooooo jooooooooo jooooooooooooooooj jj j j j j jjjj jjz 5kooooooookokoooookook ³3upoooooooooooooooo
Jun k😊
*My takeaways:*
1. Artificial Intelligence (AI) vs Machine Learning (ML) vs Deep Learning (DL) 4:36
2. What is intelligence 6:27
3. Deep learning applications 7:05
4. Unsupervised learning, reinforcement learning and supervised learning 11:04
5. Training and test, with examples 16:28
6. Classification, k-mean classifier 31:05
7. Neurons 35:35
8. Fully-connected layers 40:45
9. Activation functions, softmax 46:16
10. Loss, cost 52:12
11. Gradient descent 55:45
12. Backpropagation 1:03:30
13. Learning with gradient descent, learning rate, convergence 1:07:46
14. Data 1:19:36
15. Dimensionality reduction, Principal Components Analysis (PCA) 1:21:22
16. Evaluating accuracy: across-validation 1:27:30
17. Overfitting and underfitting, regularization (e.g. dropout) 1:48:10
18. Convolution, with examples 2:01:55
19. Autoencoder, latent variables, variational autoencoder, denoising 2:33:00
20. Reinforcement learning, Q learning 2:46:38
21. Recurrent Neural Networks 2:54:25
22. Generative adversarial networks (GANs), deep dream, style transfer 2:58:22
23. Ethics 3:20:46
That interesting
@@leixun Thanks for timestamps. Im not sure it is a takeaway
Imk
Imk
@@adiflorense1477 9ijjk9lmmk08om
Woke up to TH-cam auto playing this so I assume I absorbed this information as I slept
There is always an answer to every question that pops into my head!
Can't thank you enough 👏🏼
Quality stuff, great content and way of presenting it, one of the best tech lectures I've seen on youtube.
The concepts are so clearly explained i like that. It was really nice to watch Great content, awesome explanations, great job. the best explanation of deep learning I've seen
I've been working with this in the background. When they took break... I got such a (pleasant) surprise when he called "1 minute", I'd forgotten this lecture was still happening.
Definitely one of the best lectures on Deep Learning. Andrew's teaching style and delivery makes this otherwise hard subject so easy to understand. The concepts are so clearly explained that even the first time audience can grasp the bulk of it if not all. It was really exhilarating to watch and listen to this whole lecture and dispel any misconceptions about certain techniques that may have been misconstrued or unintentionally taught wrong by others.
I definitely learned a lot from it. Thank you sir for making this video available to all and to learn from.
Wow. Amazing. Thank you - great lecture. I never suspected that a 3+ hr. lecture could be so interesting!
expected**
@@devendraparhate4296 ⁹99⁹9999999⁹⁹9⁹⁹99999⁹99999⁹9⁹⁹99999⁹99999⁹999999999⁹⁹999⁹⁹⁹⁹9999⁹⁹⁹99999999⁹9999999999⁹99999999⁹9999999⁹999999999⁹999999⁹⁹99999999⁹99999999⁹⁹99999999⁹999999⁹9999999⁹⁹99999⁹999999⁹9999⁹9999⁹99999⁹99999⁹9999⁹⁹9999999⁹⁹99999⁹99999⁹9⁹99999⁹99999⁹⁹9999⁹99999⁹9999999⁹99999⁹999999⁹9999999⁹999⁹99999⁹9999⁹99999⁹99999⁹999999⁹999⁹9999999⁹99999⁹9999⁹999⁹999⁹9999⁹9999⁹99999⁹99999⁹999⁹⁹999⁹999⁹999⁹999⁹999999⁹999⁹9999⁹9999⁹9999⁹9999⁹9999⁹999⁹⁹9999⁹9999⁹999⁹9999⁹999999⁹99⁹9999⁹999⁹99⁹99⁹999⁹9999⁹9999⁹9999⁹⁹999⁹999⁹999⁹999⁹9999⁹9999⁹⁹9999⁹999⁹99⁹9999⁹99999⁹9999⁹9999⁹9999⁹999⁹99999⁹9999⁹99999⁹999⁹99999⁹99999⁹9999⁹9999⁹9999⁹9999999⁹99999⁹999999⁹999999⁹999999⁹999999⁹99999⁹999999⁹⁹99999⁹9999⁹99999⁹99999⁹99999⁹99999⁹9999⁹99⁹99999⁹99999⁹⁹9999⁹9999⁹99999⁹999999⁹9999⁹9999⁹99999999999999⁹9999⁹⁹999999⁹99999⁹9999⁹999999⁹999⁹99⁹99999⁹999999⁹99999⁹99999⁹9999⁹999999⁹9999⁹99999⁹999⁹999⁹99⁹999⁹999⁹9999999⁹999⁹99999999⁹9999⁹99999⁹999⁹99999⁹9999⁹9999999⁹99999⁹⁹9999999⁹999999999⁹⁹9999999999⁹9⁹9999999999⁹⁹999999999⁹99999999999⁹999999999⁹999999⁹9999999⁹⁹9999999⁹9999999⁹99999⁹99999⁹999⁹999999⁹9999⁹9999⁹999⁹99⁹999⁹99999999⁹99999⁹999999⁹999999⁹999⁹9999⁹99999999⁹9999⁹99999⁹999999⁹9999999⁹9999999⁹⁹999999⁹9999999⁹9⁹99999999⁹9999999⁹⁹999999⁹⁹999999⁹⁹999999⁹9999999⁹9999999⁹99999999⁹999999⁹99999⁹99999999⁹99999⁹999999⁹999999⁹999999⁹99999⁹9999⁹99999⁹9999⁹99999⁹9999⁹999999⁹99999⁹9999⁹999999⁹⁹99999⁹9999⁹999999999⁹99999⁹99999⁹9999⁹99⁹999⁹99999⁹999999⁹9999⁹99999⁹9999⁹9999⁹999⁹999⁹9999⁹999⁹99999⁹9999⁹999⁹9999⁹99⁹999⁹999⁹9999⁹9999⁹999⁹99⁹999⁹99⁹99⁹9999⁹9999⁹9999⁹999999⁹99⁹9999⁹99⁹999⁹99⁹9999⁹999⁹9999⁹9999⁹999999⁹99999⁹999999⁹999⁹999⁹9999⁹99999⁹9999⁹99999⁹9999⁹999⁹99999⁹99999⁹99999⁹99999⁹99999⁹99⁹999⁹9999⁹999⁹9999⁹999⁹9999⁹9999⁹9999⁹99⁹9999⁹99999⁹9999⁹9999⁹99999⁹9999⁹9999⁹999⁹99999⁹9999⁹999⁹9999⁹99999⁹99999⁹999⁹9999⁹9999⁹9999⁹999999⁹999⁹9999⁹99999⁹9999⁹99999⁹99⁹999⁹999999999999⁹9999⁹9999⁹9999⁹9999⁹999⁹99999⁹99⁹99⁹999⁹999⁹9999⁹999⁹9999⁹99999⁹999⁹99999⁹9999⁹9999⁹999999999⁹99⁹999⁹99⁹999⁹99999999⁹999⁹99999⁹99⁹999⁹9999⁹99⁹99⁹999⁹99⁹99⁹999999⁹999⁹9999⁹99⁹9999⁹99⁹99⁹99⁹999⁹9999999999⁹9⁹999⁹9⁹99⁹999⁹999⁹9999999⁹9999999⁹9999⁹9999⁹999⁹99⁹999999999⁹999⁹999⁹999999999999999⁹99⁹999⁹9999⁹9999999999⁹9999⁹999⁹9⁹99999⁹9999⁹999⁹999⁹999⁹9999⁹9999⁹999⁹9999⁹999⁹999⁹9999⁹999⁹999⁹999⁹999999999999⁹9⁹9999999999999999999⁹9⁹9⁹9999⁹99999999999999999999999999999999999999⁹999999999999⁹9999999⁹999999⁹99999999999⁹9999999999999999999⁹9⁹99⁹99999⁹9999999999999999999999999999999999999⁹99999⁹999999⁹9⁹99999999999⁹999999999999999999999999999999999999999999⁹999999999999999999999999999999999999⁹99999999999999999999999999999999⁹999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999⁹999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999dd
SIGGRAPH please don't delete these videos
Bcuz i had seen that there are no videos of last year
Please don't delete it as
Its valueable to new peoples, this lectures motivates us to study more about COMPUTER GRAPHICS
I'm sure that we new peoples will give SIGGRAPH more researchers in few years
Please support cummunity by giving them valueable lectures 😇😇😇
I legit just found video. it's legit. Hope they stay up.
Thanks for sharing the insight that they disappear - will download this one with some PP dowloader :) - BTW the guy was talking for over 3 hours without water - thats a lot! @@SolarBang
Andrew wrote a book on raytracing and that made that subject matter accessible to many. Now this and recent SigGraph Quantum Tutorial are a great contribution to science of deep learning and quantum computing!
Thank you Andrew Glassner for sharing this information, translating it, preparing it and metaphorizing it. Thank you for speaking up about ethics as well.
Woooww He just compiled my entire semester classes un one course of 3 hours, amazing. Great and Thanks for that.
i am loving this guy
Thank you so much! It helped me get a general idea about the basic ideas of deep learning.
This is pure gold. Connected so many scattered nodes for me, and added many more. Worth watching over and over.
Thank you Andrew & SIGGRAPH for sharing this excellent introductory lecture on deep learning!
informative talk and well presented thanks for uploading :)
The presentation is so explanatory and simple.
خظ..ح.كححح... .
omg he explain the meaning of regretion (and a bunch of other things that teachers sometimes don't tell) watch this lecture worth alot
Regression*
You make it sound regretful lol
AaA❤ see❤ see❤ we w WWE a❤ WWE❤aww see a sea aa❤❤ aw WWE❤ wa❤❤❤❤aa
😮Days 😊😢😊🎉😅😢😅D😢😊😊😢😢I 🎉🎉😢🎉🎉😮😅😢😊ays . 🎉😮I😊I'🎉😢😢😢😢😢EEE@😢
Excellent talk!This Course inspired me to buy the books and continue learning with this outstanding teacher.Thank you ver much
I loved your presentation. It is the best explanation of deep learning I've seen. Thank you for sharing all your hard work and knowledge.
interesting.. I fell asleep and had some cool dreams revolving around the CCNA and woke up to this.. nice!
nice way of teaching especially changes of voice and very informative but time is short to cover these topics effectively
Amazing session, you explain very well. I was struggling with Deep learning concept but after watching your video I am feeling more confident now. I would love to follow your more video regarding Machine learning.
This is the kind of deep learning explanation I have been looking for. It is simple yet deep. It is one of the best videos I found online to explain to a friend what singularity actually means and is all about. Thank you professor for sharing.
Wow that was an awesome talk. The host was very entertaining and informative to listen to.
so good, this course. way better than what I had in Uni.
excellent overall coverage, great lecture
I finally found the "Deep Learning" treasure course I was looking for for many weeks. Excellent in all aspects. While I was about to abandon learning deep learning, due to heavy use of all sorts of mathematics by other contributors, I am finally back on track to pursue this amazing subject. It's amazing how the concepts of deep learning were introduced without the need for any kind of mathematics. Well done Prof. Thank you.
Thank you/ This was great info/ Keep on creating .
Really good lecture. One of the best ones I have ever seen. Good tempo, good voice, great insight. My regards to Andrew Glassner.
Loved the convulsion explanation and the face explanation. So clear, so simple to understand. Awesome video, thank you! :)
the first time i heard about convolution, i went into convulsion
i was sleeping the whole time and i heard everything word by word thing is deeply engraved in my brain now
Great presentation: well done!..
Really awesome explanation for convolution.
I see this very helpful for analyzing every pattern
Fabulous lacture
Thank you for uploading
Very nice video and awesome explanation. Thank you so much.
This is one of those really good teachers. He really makes it look easy(ishh??)
Great stuff. Thanks a lot for sharing.
I don't know how i got here... I do not have a science background at all but this is so interesting and well-explained! Even for someone like me who has zero experience & knowledge in this field and topics!
Very nice presentation. it is worth to listen!
Great content, awesome explanations, just the right tempo and a brilliant orator. The most awesome introduction that I stumbled upon. Lots of kudos!
Im currently training an AI on context switches which is a major bottle neck in CPU latency, Ive been working on this for about a year Im getting better and better and figuring it out thats why I love these vids onnx models are the easiest to work with a compatibility
Wonderful course! Thank you.
this is a pretty great class.
Thank you! Lovely presentation
Very good course & great speaker! Thanks
Deep Learning makes us learn, relearn, and unlearn new things. This video is very very useful.
Thank you/ This was great info/ Keep on creating .
Thank you Andrew & SIGGRAPH for sharing this excellent introductory lecture on deep learning!
Woow...preaching Deep Learning
Thank you for sharing this informative video! 😺🖐
Great presentation! Thanks. Was there music in the background? ;) I haven't noticed or heard anything!
2:51:00 the discounting is actually more about introducing a notion of time-value to your rewards (+some mathematical convenience) rather than one of uncertainty.
Actually, it's both. Because a remembered event will be farther in the future, it therefore has less uncertainty to that state's reward. This becomes very relevant in the case of scarce rewards, and long plays, where even though an event is very far in to the future, the machine will still pursue it as the most optimal choice. (Delayed gratt)
Great presentation: well done!
T5,55t,t,6tt5it55,5,t7,5twT,55t,t5,,5et5,tc4d5t5tu55t,tzt6t,5r5,57,r5ttt 5,tt65t5,te,y5t,t5t5t57t,5t,t5t5t53t5t,t5t,t,,t5,6t5t4t5t5tutt7t,t,4tu5,5t47,57tu5tu64yr
Great stuff. Thanks for that
Great Presentation!
I fell asleep and when I woke up, this video was playing on my computer. The time on the video was 1:09:14 I watched it for nearly a whole 10 minutes before I realised this wasn't a video about stocks and shares.
u
Thanks a lot for uploading this video!! :-)
i kow that i passed out o the couch, when i got 3 hours into this o.o
I loved all the misc pictures added among the slides, like the socker football in the globe stand.
Is there a set of those?
k fold cross validation is not for preserving precious data....it's diving the data into traning and testing portions. For imbalalnced data, we use techniques like SMOTE...pls correct if I am wrong
Superb lecture
Very Good Lecture sir
this video helpful for us
Great presentation: well done!
Thank you so much! It helped me get a general idea about the basic ideas of deep learning.
Thank you Professor, for saving the time to browse the topics individually online
fell asleep and woke up to this
Thank you so much! It helped me get a general idea about the basic ideas of deep learning.
Great presentation: well done!
Second part 1:48:53
Interesting, I learned much
This is a gift 🎁
Starts @ 1:40
I really enjoyed the lighted hearted approach of this talk to explaining really 'deep' concepts in deep learning.
L lol l
@@babajaveedb you you have have have fun at church and get get get
@@babajaveedb you guys ruturut
Trur you
Tututrurtuturt
I woke up to this playing in the middle of the night
Ok
the ok is worse just for the record....I'm dropping a stinky half eaten tuna here @@atetraxx
Awesome,very nice friend👌👍
Brilliant talk, learned a lot!
l
was cleaning then went to go take a break i bet my pc was enjoying this video
Great.
Thank you
woke up to this video
Yo same!! Haha
same
thanks a lot!
This is a fantastic talk!
Thank you so much
This chapter is very helpful when deciding what to sell and how to price.
Thanks a lot for these great explanations, was really helpful !.
But are the comments here mostly created by a bot ?. Because most of them say exactly the same I just noticed ?.
I noticed that too. The comments seem to be created by fake accounts. By the looks of it, the algorithm shuffles sentences of existing comments to create new comments.
You Tube comments do this all the time. The Ai and algorithms target certain people and videos they want them to watch then repeat certain comments over and over to drum a message into there head.
Oh, just a hour ago I bought a book (Deep Learning, A Visual Approach: No Starch Press; Illustrated edition (June 29, 2021)) by this guy:
Starts off playing Steely Dan, good taste ;)
I keep getting woken up by this video though, it's annoying.
Brilliant, I absolutely loved the explanation of Deep Dream. It's so insanely simple, yet so insanely complex.
Thank you :)
My neuron finally understood what the neuron networks do, thanks a lot
I love how he explains everything clearly. I wonder how this material holds up after the release of large language models. Hopefully these core concepts are still relevant.
Being a nuerodyvergent person while dating a nuerodyvergent person and both having semi traumatic childhoods basically forced my mind open to the higher levels in order to stay together and I feel like a much better person ever since. This put video is the best explanation I've ever seen keep it up man.
Steely Dan and ML. A combination tough to beat.