That’s a great visual. Thanks! 
1 Like
Not mine. This one is by Brandon (Rohrer).
1 Like
GitHub - vdumoulin/conv_arithmetic: A technical report on convolution arithmetic in the context of deep learning has some pretty cool animations showing how convolutions with different strides/padding look like.
5 Likes
This is also a great and quite extensive paper:
A guide to convolution arithmetic for deep learning by Vincent Dumoulin, Francesco Visin.
1 Like
Adding to this,one the Stanford CNN lecture which explain CNN kernels and strides pretty well
2 Likes
1 Like
@bhutanisanyam1 very well said, totally agree!
Thank you so much @bhutanisanyam1 for sharing my blog!
EDIT:
I’ve managed to write blog posts because of the amazing study groups with this community. I’m learning so much every single week! So thank you for that!
3 Likes
4 Likes
As Sanyam mentioned start learning with a top-down approach, in this way you will be able to learn the practical aspects of Deep learning in the shortest span of time and with solid foundational intuitions and most of well know members of the ML community like Jeremy Howard, Zach Mueller vouch for the same. You can see from Emil Wallner’s the recent Tweet
2 Likes