Convolutions in Image Processing | Week 1, lecture 6 | MIT 18.S191 Fall 2020

What software is he using for scripting? The visualization of codes is pretty straightforward.

what is this programming language? and the ide?

Wouldn't the elemental inverse of the kernel, oh, I see

what's the live docs thing that he pulled up on the side, seems much more efficient in case I needed to know the methods of particular module instead of going over to google the documentation or stackoverflow all the time, sure I could just use dir to list out all methods of a module in python but this seems cool. Has anyone got an idea ?

Hey Grant Sanderson, you did a good job on your short circuit story about this project, you really did it. you will might reach more then 3 floor underground to reach this secret service department.

You could probably get the AI to un-blur too?

Sub spanish? :,v o mejor aprendo inglés jaja:(

I hope to continue my PhD in image processing with your team teacher .. I am from Iraq 😔

I am sure with some convolution the horrible audio quality could have been significantly increased. Despite that, and joke aside, great content and awesome visualization of what convolving functions means

Been watching your videos for quite a few years, and I always wanted you to teach me. But now that I have seen your face, i want you to teach me, daddy 🛐🛐

if he was my teacher i would come every day (he's so cute)

Kudos for excellent series.

Great lecture . I also use math phix, it is also a nice tool! :)

Thanks for this content. Really a master class brilliantly explained. I wonder if we could also see similar lectures on subjects such as rotational, divergence, gradient (relationship amongst them) and some of the properties of typical operations that interrelate them in Maxwell equations or in other disciplines like fluids, meteorology, etc making them be intuitive and easy to comprehend.

MIT must feel grateful that they got Grant.

when he change some variable the next cells that contain that variable automatically re-run, is that default behaviour of julia notebook? how can i achieve that in jupyter notebook?

2 more tutorials lessons and we can start rebuilding Photoshop

That is awesome!

After 7 minutes of watching, I felt, the way this person speaks and their intonation is so familiar. Quick search. Oh, it is the person that narrates 3b1b!

Grant Sanderson is god for me! Thank you!

I was so confused when 3b1b's face appeared...

How do you create /edit such videos

"last time" Can someone link that up?

Very clearly and nicely explained! Well done 🤩

I wish I had such a cute professor

are you code bullet ?

what does he mean in the beginning when he says "last time" if this is his first lecture in the Computational thinking series?

very wonderful clip vdo. thank you for your content is free. But i have some error when i use function plot_1d_fourier_spectrum(img), my screen show this function not define. Point of my question is "What is the Name of packet for the function plot_1d_fourier_spectrum(img)?

Thanks! U help me more than my professor of digital image processing :)

"Is it possible to reconstruct the original from the blurred image?" Yes, I did that for a living for a major NASA mission.

Basic principle is to divide the Fourier transform of the image by the Fourier transform of the kernel. Result is the Fourier transform of a good estimate of the original image.

I am pretty much new to this , can any one explain how the 3*3 convolutional matrix/filter matrix is producing the same resolution of the picture ?

Is the used notebook is jupyter?

NICE LECTURE

Aren't you the guy from 3blue1brown?

Superb explanations! Thanks for such a nice and easy to understand lecture

Are the kernels static, or are they also trained / trainable?

You mentioned examples such as the gaussian or edge detector kernels, but it seems that on Keras I dont have the possibility to customize them.

which programming language is he using??

What software is used for the animations!? Very nice lecture :)

Why a clear image needs to be blurred in this way. What is the purpose of blurring the image anyway. Is it possible to do a clear image from a blurred version of an image, why is there a reverse process required.

Love you Grant

I've one problem with Julie that is It takes forever to install Packages

Where can I find his discord channel?

Amazing channel, I’m now subscribed, and can’t wait for more content

I'm sad

Should we not talk of cross correlation instead of convolution for this kind of operation?