ATTENTION | An Image is Worth 16x16 Words | Vision Transformers (ViT) Explanation and Implementation

This video covers everything about self attention in Vision Transformer - VIT , and its implementation from scratch.
I go over all the details and explain everything happening inside attention in vision transformer in detail through visualizations and also go over how an implementation of self-attention from scratch would look like in Pytorch.

I cover Vision transformer ( VIT ) in three parts:
2. Self Attention in Vision Transformer VIT - This video

*Other Good Resources*

*TimeStamps* :
00:00 Intro
00:33 Intuition of What isAttention & Why its helpful
03:23 Inside Attention - What is Relevant
07:53 Inside Attention - Building Context Representation
08:45 Building Context Representation For All Patches
09:45 Why Multi Head Attention
11:15 Building Context Representation For Multi Head Attention
12:35 Combining Wq, Wk,Wv matrix
13:34 Shapes of Every Matrix in Attention
14:48 Implementation Parts of Attention
15:12 Pytorch Implementation for Attention in Vision Transformer VIT
18:26 Outro

Background Track - Fruits of Life by Jimena Contreras

[DrAIScience]

Amazing explanation... i did not come accross the beautiful and easy explanation of transformers that seems extremely difficult... this channel deserves millions of subscribers 🎉

[sladewinter]

Best explanation of multi-head attention i have attended to! I already had a reasonable intuition but still gathered so much more, massive respect to your work 🙏

[sebastiancavada]

Great content! This is helping a lot!! Keep it up :)

[shashankdevraj]

Sir Can you explain dual attention vision transformers (Davit)please

[sladewinter]

Would rearranging by heads before splitting into q, k, v cause any logical difference. Just means fewer lines of code, and operations, but mostly was just curious to verify as it felt same to me.

[muhammadawais]

Helping, much appreciated. Sir how about self attention in image context