I am thoroughly enjoying this class. Thank you. I think I speak for Dali when I say that the importance of a single element of an image cannot be known. Without the vast expanse of eerie other-worldly landscape, the painting clearly loses something ;)
wow - what a wonderful lecture. I felt that James' explanation was even better than Grant's - I loved this session. May be, they should allow great teachers like these to take their own time, rather than putting them on a 30-minute constraint.
This was awesome. Thank you James! When you direct your attention to the MIT students what happens? Do they ask questions? Would be awesome to learn from their doubts...doubts we might have but not know yet.
Those energy maps remind me of the "spatio-temporal graph" (I'm not sure about the proper term for this) when solving the majority classification problem using cellular automata machine. They are similar in that both go only from top to bottom (for the automata machine, this represents the direction of time evolution), and that neighborhood of impacts in each row is a 3-cell window.
24:24 Is there an off-by-one error here? seam[i] = seam[i-1] + next_element[i, seam[i-1] Shouldn't it be instead: seam[i] = seam[i-1] + next_element[i-1, seam[i-1]
@@LeiosLabs I am glad that you replied. I understand that you have a different view. Just a suggestion: kindly consider increasing the font of the code. It was too small and illegible for me, until I increased the quality to 480p; but still it was too small to look at - at least that came in the way for my learning. Also, it seemed to me, watching your right wrist, that probably you're having RSI problems. I switched to a Goldtouch keyboard and an ergonomic mouse (right handed) some years ago, and I got rid of the pain (in my wrist as well as shoulders) in time - the key was to switch in time; too late would be sorry! Stay safe and healthy, and thank you all for the great course!
@@samm9840 I cannot increase the size of the font without removing the REPL on the side. I could switch back and forth between the REPL and text editor, but that got confusing to people before. Also, no RSI issues yet. I wear the glove because I use a drawing tablet as my mouse.
Doubt 1: 20:20 shouldn't it be -1 since we should move 1 element forward. If I am wrong please tell me why I am wrong. Doubt 2: 25:16 If the function is generating a seam( a path from 100th pixcel to bottom) then why is it returning some pixcel values twice or thrice(like 271 etc...)
1. This was specifically getting the view from 1:3 into a view from -1:1, so we needed to subtract 2 2. The consecutive 271's simply mean the seam goes downward at that point. I suppose it would have been more clear if we had used CartesianIndices instead of a vector of integers.
@@TheJuliaLanguage #2 That means the pixels are increasing or decreasing index of rows in the same column or it may also be increasing or decreasing index of columns in a same row. Means moving straight down upto some instant or also moving straight right/left upto some instant.😀 Thank You for explaining.❤️❤️
Those energy maps look surprisingly artistic.
I am thoroughly enjoying this class. Thank you. I think I speak for Dali when I say that the importance of a single element of an image cannot be known. Without the vast expanse of eerie other-worldly landscape, the painting clearly loses something ;)
wow - what a wonderful lecture. I felt that James' explanation was even better than Grant's - I loved this session. May be, they should allow great teachers like these to take their own time, rather than putting them on a 30-minute constraint.
This was awesome. Thank you James!
When you direct your attention to the MIT students what happens? Do they ask questions? Would be awesome to learn from their doubts...doubts we might have but not know yet.
Those energy maps remind me of the "spatio-temporal graph" (I'm not sure about the proper term for this) when solving the majority classification problem using cellular automata machine. They are similar in that both go only from top to bottom (for the automata machine, this represents the direction of time evolution), and that neighborhood of impacts in each row is a 3-cell window.
24:24 Is there an off-by-one error here?
seam[i] = seam[i-1] + next_element[i, seam[i-1]
Shouldn't it be instead:
seam[i] = seam[i-1] + next_element[i-1, seam[i-1]
If we add a small scaling to this algortihm this will be soo nice because only removing the unwanted parts makes the other important parts dominate.
cool lecture
@15:43, should it be j==1 instead of left==1?
4:53
Cones of Dunshire. Hilarious! :)
AMAZING
22:33 I'm with James on the style choice. I don't really find "for i=xx, j=yy " easier to read
Where can I find the code for this video????
24:50 What is a cache error?
Did anyone count how many times he said "essentially"?
Wish James had used Pluto! Could you please in future?
I will not, sorry. I do not like notebooks and feel like they inhibit learning.
@@LeiosLabs I am glad that you replied. I understand that you have a different view. Just a suggestion: kindly consider increasing the font of the code. It was too small and illegible for me, until I increased the quality to 480p; but still it was too small to look at - at least that came in the way for my learning.
Also, it seemed to me, watching your right wrist, that probably you're having RSI problems. I switched to a Goldtouch keyboard and an ergonomic mouse (right handed) some years ago, and I got rid of the pain (in my wrist as well as shoulders) in time - the key was to switch in time; too late would be sorry! Stay safe and healthy, and thank you all for the great course!
@@samm9840 I cannot increase the size of the font without removing the REPL on the side. I could switch back and forth between the REPL and text editor, but that got confusing to people before.
Also, no RSI issues yet. I wear the glove because I use a drawing tablet as my mouse.
@@LeiosLabs hey, which text editor are you using?
@@rito_ghosh Vim + julia plugin
Doubt 1: 20:20 shouldn't it be -1 since we should move 1 element forward. If I am wrong please tell me why I am wrong.
Doubt 2: 25:16 If the function is generating a seam( a path from 100th pixcel to bottom) then why is it returning some pixcel values twice or thrice(like 271 etc...)
1. This was specifically getting the view from 1:3 into a view from -1:1, so we needed to subtract 2
2. The consecutive 271's simply mean the seam goes downward at that point. I suppose it would have been more clear if we had used CartesianIndices instead of a vector of integers.
@@TheJuliaLanguage
#2 That means the pixels are increasing or decreasing index of rows in the same column or it may also be increasing or decreasing index of columns in a same row. Means moving straight down upto some instant or also moving straight right/left upto some instant.😀
Thank You for explaining.❤️❤️
R u doing it in MATLAB?
It's in Julia
For the love of god,please remove this time constraint i am not from mit but i'd like to learn to at little bit slower pace