Part 2: Convolution and Cross-Correlation - G. Jensen

Поділитися
Вставка
  • Опубліковано 27 лис 2024

КОМЕНТАРІ • 114

  • @chocodil2
    @chocodil2 6 років тому +84

    this is the good side of the internets. I learned more here than 2 weeks of class

  • @BurakAlanyaloglu
    @BurakAlanyaloglu 9 місяців тому +1

    This was an excellent video. I really congratulate your willingness and knowledge. It's great to see that there are still professors who are capable of giving enjoyable real life examples to make more sense instead of going over boring stuff just as if they aim to make concepts more unclear and less attractive. Thanks again :)

  • @hexdump8590
    @hexdump8590 5 років тому +34

    Man, you did a really nice job here. At last I learned practical uses for correlation and convolution. Thanks for making it easy for us to understand.

  • @Magnify.
    @Magnify. Рік тому

    This guy has a nice, calming voice.

  • @AlexCell33
    @AlexCell33 5 років тому +1

    You're great, you speak so simply and concise, yet what you say is so valuable!

  • @boyteam10
    @boyteam10 4 роки тому +1

    Best video ever. This 15 mins video solved my 4 hours struggle.

  • @donm7906
    @donm7906 7 років тому +6

    thank you ! I learned more from this video than reading books for 3 hours

  • @andresvodopivec5950
    @andresvodopivec5950 8 років тому +1

    This is by far the best explanation for these topics. Thanks a lot.

  • @risay79
    @risay79 6 років тому +5

    Thank you so much Sir! This is by far the best combination of Mathematical and Pictorial explanation of this topic so far.

  • @darkIronline
    @darkIronline 9 років тому +8

    Finally makes more sense to me now!, Thank you

  • @uh6537
    @uh6537 7 років тому +2

    Amazing Sir! I have tried to grasp this topic for ages though books without much success. Now I got it in 15 min with your excelltnt lecture! Thanks!

  • @dakoje2951
    @dakoje2951 5 років тому +5

    Very ASMR. Thank you

  • @satheeshsimhachalam7563
    @satheeshsimhachalam7563 Рік тому

    OMG !! It is so clear now. Wonderful explanation with real examples. Thank you professor

  • @ashutoshpati7874
    @ashutoshpati7874 7 років тому +16

    Dear Prof,
    Thank you for this wonderful lecture. After lot of confusion and mathematical mesh , I finally got this video which describes , what I really wanted to visualise. Planning to learn the whole lecture series . Once again Thank you and All The Very Best. :)
    Regards,
    Ashutosh

  • @hongt1930
    @hongt1930 6 років тому

    The best convolution idea explain ever!

  • @arivd8512
    @arivd8512 8 років тому

    Thanks, Professor Jensen. The tutorial helps a lot for starters. A lucid explanation.

  • @akshatjain07065
    @akshatjain07065 7 років тому +3

    amazing. I understood more than I did in whole week.

  • @siddharthrawat7205
    @siddharthrawat7205 9 років тому +2

    why don't we have more of good professors like you.

  • @jonathanlister5644
    @jonathanlister5644 7 місяців тому

    Great clarity! Thank you.

  • @mehedihassan8649
    @mehedihassan8649 6 років тому

    I wanted to push the like button for so many times!!

  • @EngineerAnandu
    @EngineerAnandu 3 місяці тому +1

    Veey good.

  • @titanh-odc6742
    @titanh-odc6742 2 роки тому

    You are the man!!!

  • @danielku4689
    @danielku4689 6 років тому +1

    Gold lecture. Perfection!

  • @marwanal-yoonus280
    @marwanal-yoonus280 4 роки тому

    Thank you very much for your good illustrations.

  • @ottmanpark
    @ottmanpark 6 років тому

    This is best lecture to help understand convolution and cross-correlation:)

  • @sepijortikka
    @sepijortikka 5 років тому

    That Cross-Correaltion plot looks like a cloud, interesting.

  • @cpt_4meric4
    @cpt_4meric4 4 роки тому

    Thank you for explaining this so well. My Professor couldn't.

  • @TheOldProgramming
    @TheOldProgramming 4 роки тому

    This is beautiful. Very well explained. Thanks and looking forward for more lessons on Computer Vision :)

  • @afonsomendes92
    @afonsomendes92 3 роки тому

    please add the previous lessons to the description!

  • @sanskarshrivastava5193
    @sanskarshrivastava5193 3 роки тому

    Damn , this is beautiful !

  • @thespiritualsabha7162
    @thespiritualsabha7162 6 років тому

    superb!!! i got it all with no confusion. thanks

  • @redxxfour
    @redxxfour 6 років тому

    The examples made it very easy to understand. Thank you

  • @tildebengtsson865
    @tildebengtsson865 4 роки тому

    Thank you for a pedagogic video!

  • @ismailsarwar733
    @ismailsarwar733 4 роки тому +1

    I think, when we use convolution theorem on the cross correlation then either f or h function should be conjugated before multiplying..

  • @harirao12345
    @harirao12345 6 років тому

    Outstanding! Thank you!

  • @rezasamangouei6979
    @rezasamangouei6979 5 років тому

    awesome description. thanks a lot.

  • @itsmerahul108
    @itsmerahul108 7 років тому +5

    Amazing..

  • @sachin.george96
    @sachin.george96 6 років тому +1

    Thank you sir .. i spent years trying to figure this out ..

  • @earthlover1871
    @earthlover1871 6 років тому +5

    very great video but i was wondering why both has the same equation in fourier domain?

    • @rustyrusky
      @rustyrusky 5 років тому

      The Fourier transform of a flipped function (i.e. f(-x)) is the complex conjugate of the Fourier transform of the original function f(x). The convolution reduces to a product in the Fourier domain and the cross-correlation reduces to a product with one function being complex conjugated.

  • @srest0173
    @srest0173 9 років тому

    awesome videos. thanks for these

  • @Chibiwobot
    @Chibiwobot 3 роки тому

    Thank you very much professor.

  • @TylerMatthewHarris
    @TylerMatthewHarris 7 років тому

    Thank you so much. You finally made it click for me

  • @thekatyperrymemechannel2122
    @thekatyperrymemechannel2122 3 роки тому

    How could image values be negative though? Aren't they always 0-255, or 0-1?

  • @rendianwar0664
    @rendianwar0664 3 роки тому

    fantastic! thanks.

  • @newjaa122
    @newjaa122 9 років тому

    Thank you very much. I'm clear about convolution and correlation

  • @anasbahi8371
    @anasbahi8371 3 роки тому

    thank you very much

  • @wenbofeng4516
    @wenbofeng4516 3 роки тому

    Make so much sense !

  • @szhavel
    @szhavel Рік тому

    what should we do if we have images in 0-255 values? we need to subtract mean value of probe and original image to get negative values?

  • @theblacktechexperience
    @theblacktechexperience 4 роки тому

    My only question is if a pixel has value of 0-255 (via RGB), then how can the multiplication of the first and second image pixel be a negative number. What did I miss?

    • @lukasd75
      @lukasd75 Рік тому

      I have a different question: What if my pattern concerns low, but positive numbers... cross correlation would be higher for places with high positive values in the test image. I guess, I am missing something, too.

  • @rozikrazimator
    @rozikrazimator 6 років тому

    Such a good video

  • @benoitv9463
    @benoitv9463 7 років тому +1

    Awesome explanation, thanks!

  • @weirdsciencetv4999
    @weirdsciencetv4999 Рік тому

    Is there a way to make cross correlation insensitive to rotation and scale?

  • @yevgeniymen6160
    @yevgeniymen6160 4 роки тому

    wow, clearly explained. Thank you!

  • @MeKaashu
    @MeKaashu 3 роки тому

    Does Sheldon Cooper still bother all of you at Caltech?

  • @bhimeshjetty7092
    @bhimeshjetty7092 6 років тому

    Thank you so much sir for clarifying with practical examples.

  • @SciHeartJourney
    @SciHeartJourney 4 роки тому

    Thank you!

  • @enen2777
    @enen2777 2 роки тому

    Thank you, Sir. Wonderful explanation.

  • @waroon_khaloon
    @waroon_khaloon Рік тому

    Shouldn't the cross-correlation function c = IFT{FT{f} x FT*(h)}, where * represents the conjugate of the function?

  • @arashboustani38
    @arashboustani38 9 років тому

    superb...

  • @bastienmoliere8325
    @bastienmoliere8325 9 років тому

    Thank you sooooo much !
    Amazing

  • @aimanyounis8387
    @aimanyounis8387 3 роки тому

    what do you mean when we do convolution one of the function flips? I did'nt get that.

  • @shangyingao7553
    @shangyingao7553 4 роки тому

    difference between convolution and cross-correlation is at 12:01

  • @angkhoapham8625
    @angkhoapham8625 9 років тому

    Can you please tell in which book should I read to dig deep into these issues?

  • @도미솔-b2w
    @도미솔-b2w 6 років тому +1

    Thank you for your lecture

  • @sam-zy2dn
    @sam-zy2dn 5 років тому

    At 6:38 he uses frequency domain to calculate convolution. But he uses the same formula at 12:45 to use it for correlation. why?

  • @ivanchan9710
    @ivanchan9710 6 років тому

    Wish I could give 1000 likes to this video

  • @NskLabs
    @NskLabs 3 роки тому

    Now, the stupid thing about this video is no matter how many times I click on thumbs up it only counts as one.

  • @tommyyhli
    @tommyyhli 8 років тому

    Thank you so much

  • @nguyenanhminhxd
    @nguyenanhminhxd 7 років тому

    Thankyou Professor!

  • @Rock57811
    @Rock57811 5 років тому

    Thanks so much!

  • @tsc411
    @tsc411 4 роки тому

    The Best

  • @aoihyoudou
    @aoihyoudou 3 роки тому

    can someone help, so what exactly is the difference between the two?

  • @laxmisuresh
    @laxmisuresh 4 роки тому

    Very nice and useful lecture. Thanks sir.

  • @elrik1928
    @elrik1928 4 роки тому

    What in the actual f am i doing here at 3 am

  • @SF-fb6lv
    @SF-fb6lv 5 років тому

    5:18: Now you can jump into modulation transfer function...

  • @akhilmalik666
    @akhilmalik666 8 років тому +1

    so nice

  • @karthikmurthy2511
    @karthikmurthy2511 6 років тому

    Thanks a lot sir for these lectures.

  • @ibrahimahmethan586
    @ibrahimahmethan586 5 років тому

    god bless u . helpful

  • @mosestrosin
    @mosestrosin 6 років тому

    Thanks a lot! It's realy usefull for me!

  • @HyunjongNam
    @HyunjongNam 6 років тому

    two thumbs up!

  • @hypno5645
    @hypno5645 8 років тому

    Hello
    I don't understand around 11:30 why should a pixel value be negative ?? Isn't it supposed to be between 0 and 255 ? And so i don't understant this part. Help me please

    • @willboler830
      @willboler830 7 років тому

      The data doesn't necessarily need to be restricted to image data or 8-bit values. Images are just an intuitive example that help us understand convolutions and cross correlations.

  • @카멜-z5k
    @카멜-z5k 5 років тому

    Nice explanation. Really Thank you.

  • @AbdAlbaryTaraqji
    @AbdAlbaryTaraqji 3 роки тому

    Thank you

  • @tgnana2
    @tgnana2 9 років тому

    Wonderful lecture. I just don't understand how come, the equations for both correlation and convolution are same. (At 12:30)

    • @Tordek
      @Tordek 9 років тому

      +Gnana Thedchana Moorthy They're similar, but the critical difference is that in Convolution, you use h(i-x, j-y), and in Cross-Correlation you use h(i+x, j+y).

  • @yiyou6529
    @yiyou6529 8 років тому +6

    also for the fourier transform expression for cross correlation, you missed the complex conjugate of the f(x). the key difference between convolution and cross correlation is the space of integration. convolution integrates in displacement space while cross cottelation is in independent variable space. you are misleading people, i would suggest you to remove and revise the video.

    • @madteamaster
      @madteamaster 8 років тому +1

      I agree, I was very confused until I noticed the complex conjugate part on wikipedia!

    • @madteamaster
      @madteamaster 8 років тому +1

      hmm, actually the complex conjugate part did not really help, I still don't really understand how to use fft to do cross correlation in practice...

    • @yiyou6529
      @yiyou6529 8 років тому +3

      madteamaster Cab(v) = F*(v) ×G(v) . note that everything here is in Fourier space. then the ifft of cab(v) will give you the Cab(τ). I don't think the math here is a problem. but when you do this, assigning τ values will be a big problem.

    • @madteamaster
      @madteamaster 8 років тому +2

      Thanks, I understand now.
      (I also had issues related to the cyclic nature of the fft, which I just solved with padding.)

    • @c.h.1073
      @c.h.1073 6 років тому

      @@madteamaster Can you elaborate how you used padding to solve your problem?

  • @greenhills112
    @greenhills112 7 років тому

    very nice

  • @jhesuslegarda4026
    @jhesuslegarda4026 8 років тому

    Can you explain better that you said in 4:45 min? Thank You, nice duck lattice hahhaha

  • @quiteSimple24
    @quiteSimple24 5 років тому

    Thank you :D

  • @beevees1636
    @beevees1636 4 роки тому

    Now I understand gaussian blur from Photoshop hahahaha

  • @nermeenalriz1236
    @nermeenalriz1236 6 років тому

    thanks a lot the was so good

  • @yiyou6529
    @yiyou6529 8 років тому

    the independent variable you used for convolution seems to be incorrect. the integral of convolution is di and dj, while maintaining the independ variable of the output function and input function the same. g(x)=∫f(x)⊗h(i-x)di

    • @sonimohapatra9254
      @sonimohapatra9254 8 років тому

      That would actually make sense. Thanks

    • @abdelrahmangamalmahdy
      @abdelrahmangamalmahdy 7 років тому

      Yi You that is incorrect.

    • @abdelrahmangamalmahdy
      @abdelrahmangamalmahdy 7 років тому

      the actual variable is i .. x is just a dummy variable that's gonna get integrated out

    • @yiyou6529
      @yiyou6529 7 років тому

      Truth Seeker please check Powell and Hieftje, 1978, correlation based file searching.
      And Isao Noda, 1993, 2d-correlation spectroscopy.
      No need to argue. I have given three talks in international conferences already.

    • @abdelrahmangamalmahdy
      @abdelrahmangamalmahdy 7 років тому

      Yi You
      I'm not here to argue. I'm here to correct you.
      here we're talking about convolution not correlation.
      the correct form is just as he wrote. look at what you wrote once again and try to find out your mistake yourself.

  • @4141ca
    @4141ca 8 років тому

    tooo good :)

  • @luisperdigao6204
    @luisperdigao6204 4 роки тому

    Wrong. 12:43. The cross-correlation 'theorem' should have one of the terms being the complex conjugate. c = F-1 [ F(f)* . F(h) ] with * representing the complex conjugate. As it is presented here is the same formula as the convolution, which makes no sense.

  • @AvantGrade
    @AvantGrade 4 роки тому

    very helpful

  • @jessehansen6441
    @jessehansen6441 8 років тому

    why is the cross-correlation readout (top right @ 12mins) a sharp (curved) peak rather than a square shaped peak? The curved peak implies that the center of the image matches better than the edges of the image. When comparing, it should go from low/zero on almost every position then suddenly "snap" into place and every single pixel in the small square should match with the large square...

    • @Qxismylife
      @Qxismylife 8 років тому

      I am sure it is rather representing the coordinate of the entire probe image (where the probe image fits the best) so it will go from (0,0) to (10000,10000) and finds that (3000,2000) matches the best, since there are 10000*10000 of different possible positions for the probe image (10000 pixles* 10000pixlea)

  • @adkfunk
    @adkfunk 2 роки тому

    Thank you!