Vision Transformer for Image Classification

[@UzzalPodder]

Great Explanation with detailed notations. Most of the videos found in the KZbin were some kind of oral explanation. But this kind of symbolic notation is very helpful for garbing the real picture, specially if anyone want to re-implement or add new idea with it. Thank you so much. Please continuing helping us by making these kind of videos for us.

[@ai_lite]

great expalation! Good for you! Don't stop giving ML guides!

[@mmpattnaik97]

Can't stress enough on how easy to understand you made it

[@drakehinst271]

These are some of the best, hands-on and simple explanations I've seen in a while on a new CS method. Straight to the point with no superfluous details, and at a pace that let me consider and visualize each step in my mind without having to constantly pause or rewind the video. Thanks a lot for your amazing work! :)

[@thecheekychinaman6713]

The best ViT explanation available. Also key to understand this for understanding Dino and Dino V2

[@drelvenkee1885]

The best video so far. The animation is easy to follow and the explaination is very straight forward.

[@adityapillai3091]

Clear, concise, and overall easy to understand for a newbie like me. Thanks!

[@valentinfontanger4962]

Amazing, I am in a rush to implement vision transformer as an assignement, and this saved me so much time !

[@randomperson5303]

lol , same

[@sheikhshafayat6984]

Man, you made my day! These lectures were golden. I hope you continue to make more of these

[@aimeroundiaye1378]

Amazing video. It helped me to really understand the vision transformers. Thanks a lot.

[@thepresistence5935]

15 minutes of heaven 🌿. Thanks a lot understood clearly!

[@Peiying-h4m]

Best ViT explanation ever!!!!!!

[@soumyajitdatta9203]

Thank you. Best ViT video I found.

[@vladi21k]

Very good explanation, better that many other videos on KZbin, thank you!

[@MonaJalal]

This was a great video. Thanks for your time producing great content.

[@arash_mehrabi]

Thank you for your Attention Models playlist. Well explained.

[@DerekChiach]

Thank you, your video is way underrated. Keep it up!

[@wengxiaoxiong666]

good video ,what a splendid presentation , wang shusen yyds.

[@MenTaLLyMenTaL]

@9:30 Why do we discard c1... cn and use only c0? How is it that all the necessary information from the image gets collected & preserved in c0? Thanks

[@abhinavgarg5611]

Hey, did you get answer to your question?

[@swishgtv7827]

This reminds me of Encarta encyclopedia clips when I was a kid lol! Good job mate!

[@ronalkobi4356]

Wonderful explanation!👏

[@nova2577]

If we ignore output c1 ... cn, what c1 ... cn represent then?

[@nehalkalita]

Nicely explained. Appreciate your efforts.

[@NisseOhlsen]

Very nice job, Shusen, thanks!

[@rajgothi2633]

You have explained ViT in simple words. Thanks

[@hongkyulee9724]

Thank you for the clear explanation!!☺

[@lionhuang9209]

Very clear, thanks for your work.

[@sehaba9531]

Thank you so much for this amazing presentation. You have a very clear explanation, I have learnt so much. I will definitely watch your Attention models playlist.

[@muhammadfaseeh5810]

Awesome Explanation. Thank you

[@mmazher5826]

Excellent explanation 👌

[@chawkinasrallah7269]

The class token 0 is in the embed dim, does that mean we should add a linear layer from embed to number of classes before the softmax for the classification?

[@xXMaDGaMeR]

amazing precise explanation

[@ASdASd-kr1ft]

Nice video!!, Just a question what is the argue behind to rid of the vectors c1 to cn, and just remain with c0? Thanks

[@DrAIScience]

How data A is trained? I mean what is the loss function? Is it only using encoder or both e/decoder?

[@aryanmobiny7340]

Amazing video. Please do one for Swin Transformers if possible. Thanks alot

[@t.pranav2834]

Awesome explanation man thanks a tonne!!!

[@parmanandchauhan6182]

Great Explanation.Thanqu

[@user-wr4yl7tx3w]

In the job market, do data scientists use transformers?

[@tallwaters9708]

Brilliant explanation, thank you.

[@ervinperetz5973]

This is a great explanation video. One nit : you are misusing the term 'dimension'. If a classification vector is linear with 8 values, that's not '8-dimensional' -- it is a 1-dimensional vector with 8 values.

[@deeplearn6584]

Very good explanation subscribed!

[@jidd32]

Brilliant. Thanks a million

[@boemioofworld]

thank you so much for the clear explanation

[@saeedataei269]

great video. thanks. could u plz explain swin transformer too?

[@mariamwaleed2132]

really great explaination , thankyou

[@bbss8758]

Can you explain yhis paper please “your classifier is secretly an energy based model and you should treat it like one “ i want understand these energy based model

[@medomed1105]

Great explanation

[@BeytullahAhmetKINDAN]

that was educational!

[@sudhakartummala4701]

Wonderful talk

[@ThamizhanDaa1]

WHY is the transformer requiring so many images to train?? and why is resnet not becoming better with ore training vs ViT?

[@ansharora3248]

Great explanation :)

[@zeweichu550]

great video!

[@DrAhmedShahin_707]

The simplest and more interesting explanation, Many Thanks. I am asking about object detection models, did you explain it before?

[@fedegonzal]

Super clear explanation! Thanks! I want to understand how attention is applied to the images. I mean, using cnn you can "see" where the neural network is focusing, but with transformers?

[@ME-mp3ne]

Really good, thx.

[@ogsconnect1312]

Good job! Thanks

[@DungPham-ai]

Amazing video. It helped me to really understand the vision transformers. Thanks a lot. But i have a question why we only use token cls for classifier .

[@NeketShark]

Looks like due to attention layers cls token is able to extract all the data it needs for a good classification from other tokens. Using all tokens for classification would just unnecessarily increase computation.

[@Darkev77]

@@NeketShark that’s a good answer. At 9:40, any idea how a softmax function was able to increase (or decrease) the dimension of vector “c” into “p”? I thought softmax would only change the entries of a vector, not its dimensions

[@NeketShark]

@@Darkev77 I think it first goes through a linear layer which then goes through a softmax, so its the linear layer that changes the dimention. In the video this info were probably ommited for simplification.

[@shamsarfeen2729]

If you remove the positional encoding step, the whole thing is almost equivalent to a CNN, right? I mean those dense layers are just as filters of a CNN.

[@parveenkaur2747]

Very good explanation! Can you please explain how we can fine tune these models to our dataset. Is it possible on our local computer

[@ShusenWangEng]

Unfortunately, no. Google has TPU clusters. The amount of computation is insane.

[@parveenkaur2747]

@@ShusenWangEng Actually I have my project proposal due today.. I was proposing this on the dataset of FOOD-101 it has 101000 images So it can’t be done? What size dataset can we train on our local PC

[@parveenkaur2747]

Can you please reply? Stuck at the moment.. Thanks

[@ShusenWangEng]

@@parveenkaur2747 If your dataset is very different from ImageNet, Google's pretrained model may not transfer well to your problem. The performance can be bad.

[@palyashuk42]

Why do the authors evaluate and compare their results with the old ResNet architecure? Why not to use EfficientNets for comparison? Looks like not the best result...

[@ShusenWangEng]

ResNet is a family of CNNs. Many tricks are applied to make ResNet work better. The reported are indeed the best accuracies that CNNs can achieve.

[@sevovo]

CNN on images + positional info = Transformers for images

[@yinghaohu8784]

1) you mentioned pretrain model, it uses large scale dataset, and then using a smaller dataset for finetuning. Does it mean, they c0 is almost the same, except the last layer softmax will be adjusted based on the class_num ? and then train on fine-tuning dataset ? Or there're other different settings ? 2)Another doubt for me is, there's completely no mask in ViT, right? since it is from MLM ... um ...

[@swishgtv7827]

The concept has similarities to TCP protocol in terms of segmentation and positional encoding. 😅😅😅

[@mahmoudtarek6859]

great

[@st-hs2ve]

Great great great

[@seakan6835]

其实我觉得up主说中文更好🥰🤣

[@boyang6105]

也有中文版的（ kzbin.info/www/bejne/eJPdgI1via2ln7s ），不同的语言有不同的听众

[@顾小杰]

👏

[@mahdiyehbasereh]

That was great and helpful 🤌🏻

[@randomperson5303]

Not All Heroes Wear Capes

[@yuan6950]

这英语也是醉了

[@kutilkol]

this is supposed to be english?

[@tianbaoxie2324]

Very clear, thanks for your work.

[@Raulvic]

Thank you for the clear explanation