Moacir Antonelli Ponti
Moacir Antonelli Ponti
  • 166
  • 173 962

วีดีโอ

scc0270 09 - 1 - Arquitetura Transformer
มุมมอง 7522 ปีที่แล้ว
scc0270 09 - 1 - Arquitetura Transformer
scc0270 06 - 2 - Autoencoder em pytorch
มุมมอง 3572 ปีที่แล้ว
scc0270 06 - 2 - Autoencoder em pytorch
SCC0270 - Redes Neurais 06 - 1 - Autoencoders
มุมมอง 5362 ปีที่แล้ว
SCC0270 - Redes Neurais 06 - 1 - Autoencoders
SCC0270 - Redes Neurais 04 - 5 - pytorch custom Dataset
มุมมอง 4202 ปีที่แล้ว
SCC0270 - Redes Neurais 04 - 5 - pytorch custom Dataset
SCC0270 - Redes Neurais 04 - 4 - pytorch introdução à Dataset e DataLoader
มุมมอง 3322 ปีที่แล้ว
SCC0270 - Redes Neurais 04 - 4 - pytorch introdução à Dataset e DataLoader
SCC0270 - Redes Neurais 03 - 3 - Projeto de Redes Neurais / pytorch: camadas, gradiente e otimização
มุมมอง 3902 ปีที่แล้ว
SCC0270 - Redes Neurais 03 - 3 - Projeto de Redes Neurais / pytorch: camadas, gradiente e otimização
SCC0270 - Redes Neurais 03 - 2 - Bonus: autograd pytorch
มุมมอง 1752 ปีที่แล้ว
SCC0270 - Redes Neurais 03 - 2 - Bonus: autograd pytorch
SCC0270 - Redes Neurais 03 - Implementação Perceptron Regressor em pytorch
มุมมอง 4092 ปีที่แล้ว
SCC0270 - Redes Neurais 03 - Implementação Perceptron Regressor em pytorch
SCC0270 - Redes Neurais 02 - 3 - Introdução aos Multilayer Perceptrons
มุมมอง 5902 ปีที่แล้ว
SCC0270 - Redes Neurais 02 - 3 - Introdução aos Multilayer Perceptrons
SCC0270 - Redes Neurais 02 - 2 - Intuição de modelo em pytorch
มุมมอง 4542 ปีที่แล้ว
SCC0270 - Redes Neurais 02 - 2 - Intuição de modelo em pytorch
SCC0270 - Redes Neurais 02 - 1 - Modelos lineares e Perceptron
มุมมอง 5882 ปีที่แล้ว
SCC0270 - Redes Neurais 02 - 1 - Modelos lineares e Perceptron
SCC0270 - Redes Neurais 01 - Introducao pytorch
มุมมอง 8042 ปีที่แล้ว
SCC0270 - Redes Neurais 01 - Introducao pytorch
IPE em Python (5) 3 - Estrutura de repetição while
มุมมอง 2983 ปีที่แล้ว
IPE em Python (5) 3 - Estrutura de repetição while
IPE em Python (5) 2 - Repetição aninhada com for
มุมมอง 2443 ปีที่แล้ว
IPE em Python (5) 2 - Repetição aninhada com for
IPE em Python (5) 1 - Estrutura de iteração/repeticão for
มุมมอง 2553 ปีที่แล้ว
IPE em Python (5) 1 - Estrutura de iteração/repeticão for
IPE em Python (4) 4 - Fazendo mais coisas com Listas e Tuplas
มุมมอง 1903 ปีที่แล้ว
IPE em Python (4) 4 - Fazendo mais coisas com Listas e Tuplas
IPE em Python (4) 3 - Tuple
มุมมอง 2013 ปีที่แล้ว
IPE em Python (4) 3 - Tuple
IPE em Python (4) 2 - List
มุมมอง 2553 ปีที่แล้ว
IPE em Python (4) 2 - List
IPE em Python (4) 1 - String: cadeia de caracteres
มุมมอง 5373 ปีที่แล้ว
IPE em Python (4) 1 - String: cadeia de caracteres
IPE em Python (3) 3 - Estrutura Condicional ELSE e ELIF
มุมมอง 1563 ปีที่แล้ว
IPE em Python (3) 3 - Estrutura Condicional ELSE e ELIF
IPE em Python (3) 2 - Estrutura Condicional IF
มุมมอง 1533 ปีที่แล้ว
IPE em Python (3) 2 - Estrutura Condicional IF
IPE Python (1) 3 - Instalando e utilizando Jupyter Notebook no Windows (feat. Thalisson Nobre)
มุมมอง 2113 ปีที่แล้ว
IPE Python (1) 3 - Instalando e utilizando Jupyter Notebook no Windows (feat. Thalisson Nobre)
IPE em Python (2) 3 - Entrada e saída: input() e print()
มุมมอง 2893 ปีที่แล้ว
IPE em Python (2) 3 - Entrada e saída: input() e print()
IPE em Python (2) 2 - Expressões aritméticas e lógicas
มุมมอง 1733 ปีที่แล้ว
IPE em Python (2) 2 - Expressões aritméticas e lógicas
IPE em Python (2) 1 - Memória principal: tipos de dados e variáveis
มุมมอง 2143 ปีที่แล้ว
IPE em Python (2) 1 - Memória principal: tipos de dados e variáveis
IPE em Python (1) 2 - Python como linguagem de programação
มุมมอง 2503 ปีที่แล้ว
IPE em Python (1) 2 - Python como linguagem de programação
IPE em Python (1) 1 - Algoritmos e Computação
มุมมอง 5913 ปีที่แล้ว
IPE em Python (1) 1 - Algoritmos e Computação
[Talk at UCL] Improving transfer learning with self-supervision and multiple feature alignment
มุมมอง 2193 ปีที่แล้ว
[Talk at UCL] Improving transfer learning with self-supervision and multiple feature alignment
ICC2 (4) 3 - Hashing e resolução de colisões por sondagem linear: com divisão e multiplicação
มุมมอง 1.5K3 ปีที่แล้ว
ICC2 (4) 3 - Hashing e resolução de colisões por sondagem linear: com divisão e multiplicação

ความคิดเห็น

  • @claudiorodolfosousadeolive5169
    @claudiorodolfosousadeolive5169 หลายเดือนก่อน

    Poderia ter somente ajustado as fórmula. Sem precisar desperdiçar a posição zero.

  • @StoriesFromWeb13
    @StoriesFromWeb13 2 หลายเดือนก่อน

    thank u sir very helpful

  • @Sarmoung-Biblioteca
    @Sarmoung-Biblioteca 4 หลายเดือนก่อน

    Muito bem explicado !

  • @shivamsingh-ww5on
    @shivamsingh-ww5on 4 หลายเดือนก่อน

    Excellent!, explanation

  • @purple3036
    @purple3036 6 หลายเดือนก่อน

    Luke Smith?

  • @radityawaliulu
    @radityawaliulu 6 หลายเดือนก่อน

    Thanks god I got one.. I still looking for mathmatic equation formula in python for morphology (erosion, dilation, etc)

  • @Samue361
    @Samue361 7 หลายเดือนก่อน

    obrigado !!!

  • @nishchaykumar6942
    @nishchaykumar6942 7 หลายเดือนก่อน

    1:00 how tf are we looking at the pixel to the right hand side of the pixel whom we are comparing with? its 0x1 so 0 rows and one row above. so it should be the pixel directly over the head of the main test pixel right?

    • @MoacirAntonelliPontiUSP
      @MoacirAntonelliPontiUSP 7 หลายเดือนก่อน

      it represents a shift in (row,column), therefore 0,1 means we are comparing the current pixel (at each position x,y), with the pixel x+0, y+1, therefore it would be the pixel on the right hand side for the person that it is looking at the matrix.

    • @jewel9458
      @jewel9458 17 วันที่ผ่านมา

      i also had that doubt then i figured it out, in digital image processing the matrix representation starts from top left corner and is not like a conventional graph axes, i.e., in the given figure (0,5) would be 0 according to traditional graph but using the image processing it would be (0,0) hence you have to take top left corner as (0,0) basically tilt your head right and then use the traditional graph method.

  • @naludany
    @naludany 7 หลายเดือนก่อน

    como fazer a contagem com dois for, um dentro do outro? 1. soma = 0 2. para (i = 1, i <= n, i = i + 1) faça 3. para (j = 1, i <= m, j = j + 1) faça 4. soma = soma + vetor1[i] + vetor2[j] 5. devolve soma

    • @Artoziun
      @Artoziun 5 หลายเดือนก่อน

      soma = 0 é constante => O(1) primeiro loop itera sobre n elementos => O(n) segundo loop itera sobre m elementos => O(m) soma agora é atribuida novamente, recebendo o próprio valor, mais o valor do vetor1 em dada posição e o valor do vetor2 em dada posição => O(1) + O(1) + O(1) = O(3) devolve soma retorna o valor da soma => O(1) constantes podem ser ignoradas (questão assintótica) sobra O(n) e O(m), o final seria O(n × m)

  • @rkaspar92
    @rkaspar92 8 หลายเดือนก่อน

    this was super helpful for me - thank you!

  • @lidiaperside1120
    @lidiaperside1120 9 หลายเดือนก่อน

    pra cada quicksort se usa uma fórmula geral ou essa é desse problema?

  • @ramonfernandesviana6845
    @ramonfernandesviana6845 9 หลายเดือนก่อน

    Só o que posso dizer aqui é obrigado. Que aula top, parabéns professor. Espero um dia poder ter essa naturalidade para explicar assuntos da minha área.

  • @drilonaliu150
    @drilonaliu150 10 หลายเดือนก่อน

    3:30 should be A - B = { x | x e A AND x not in B}

  • @taygarayanne9692
    @taygarayanne9692 10 หลายเดือนก่อน

    muito obrigada pelo video! foi calmo e explicou cada detalhe, me ajudando muito a compreender. 🙌

  • @diogenesalves2007
    @diogenesalves2007 ปีที่แล้ว

    muito boa a explicação

  • @gixiul88
    @gixiul88 ปีที่แล้ว

    I didn't understand what Is img_t and why we set It to 1 at the begininng

  • @Sahrul-sd5mh
    @Sahrul-sd5mh ปีที่แล้ว

    sir, why the photo called in the bag of features file is not in the images folder

  • @travelwithus3732
    @travelwithus3732 ปีที่แล้ว

    Thanks for the detailed information.

  • @alexi_space
    @alexi_space ปีที่แล้ว

    nice content

  • @jansu3127
    @jansu3127 ปีที่แล้ว

    I compared the result with numpy.fft.fft(f), and it gives a different result. So, I assume this is WRONG. Can someone know why? The code I transcribed, just in case I typed something wrong, although I checked many times... def DFT1D(f): """ DIFFERENT result than numpy.fft.fft(f)""" F = np.zeros(f.shape, dtype=np.complex64) n = f.shape[0] for u in np.arange(n): for x in np.arange(n): F[u] += f[x] * np.exp( (-1j * 2 * np.pi * u * x) / n ) return F/np.sqrt(n) Same with the DFT2D: def DFT2D(f): """ different result than np.fft.fft2(f)""" F = np.zeros(f.shape, dtype=np.complex64) m, n = f.shape[0:2] x = np.arange(n) y = np.arange(m) for u in np.arange(n): for v in np.arange(m): for x in np.arange(n): for y in np.arange(m): F[u, v] += f[x, y] * np.exp( (-1j * 2 * np.pi) * (((u * x)/n) + ((v * y)/m)) ) return F/(n*m)

    • @MoacirAntonelliPontiUSP
      @MoacirAntonelliPontiUSP ปีที่แล้ว

      Hello Jan, maybe the normalization is different. See that I divided by np.sqrt(n) in DFT and by (n*m) in DFT2D. I'm not sure how numpy's fft is normalized, but if you find out you can just replace it in the code by the same denominator and try again

  • @mauricioantonelli4145
    @mauricioantonelli4145 ปีที่แล้ว

    .

  • @mauricioantonelli4145
    @mauricioantonelli4145 ปีที่แล้ว

    .

  • @japinhastuperker1203
    @japinhastuperker1203 2 ปีที่แล้ว

    Show!

  • @agustli2851
    @agustli2851 2 ปีที่แล้ว

    I use the recursion to accomplish region growing, but the depth is beyond the max

  • @reginaldopereiradasilvafil9276
    @reginaldopereiradasilvafil9276 2 ปีที่แล้ว

    Oi, excelente vídeo, tentei aplicar o código para input como sendo matriz de array e me gerou o seguinte erro: only length-1 arrays can be converted to Python scalars, vc poderia dar uma dica de como adaptar o código ?

  • @Lakke_Edu
    @Lakke_Edu 2 ปีที่แล้ว

    Massa! Tirou minhas dúvidas, ajudou bastante. Faz um vídeo ensinando a modificar alguma informação que já esteja no arquivo binário

  • @jamespaz4103
    @jamespaz4103 2 ปีที่แล้ว

    Hello Good day, I am a college student and would liked to commend your program, video, documentation about this Image Restoration filter. This is very useful to our ongoing research, primarily the usage of gaussian filter function to convolve/blur an image. We also used and included it in our research references, and rest assured proper credits will still be given. Thank you.

  • @mauricioantonelli4145
    @mauricioantonelli4145 2 ปีที่แล้ว

    Obrigado pela aula!

  • @pedrolobo9835
    @pedrolobo9835 2 ปีที่แล้ว

    Muito bom o vídeo. Obrigado por compartilhar esse conteúdo!

  • @mauricioantonelli4145
    @mauricioantonelli4145 2 ปีที่แล้ว

    Valeu!

  • @mauricioantonelli4145
    @mauricioantonelli4145 2 ปีที่แล้ว

    obrigado pela aula e pelo ASMR de teclado!

  • @symmercool7672
    @symmercool7672 2 ปีที่แล้ว

    At 5:37 how did you normalise the GLCM matrix?

    • @symmercool7672
      @symmercool7672 2 ปีที่แล้ว

      I got it. divide each element by sum of all elements. i.e., (2+2+1+4+5+2+4 =20)

  • @PortesNew
    @PortesNew 2 ปีที่แล้ว

    Professor, tudo bem? Vai disponibilizar aulas por aqui ?? Nossa, espero que sim, seria excelente! Rsrs

  • @andersonbelgamo1142
    @andersonbelgamo1142 2 ปีที่แล้ว

    Excelente material. Parabéns!!

  • @Vapavanelli
    @Vapavanelli 2 ปีที่แล้ว

    Você é genial em sua aula !!

  • @ruzthilank9846
    @ruzthilank9846 2 ปีที่แล้ว

    thx ! good video !

  • @litevenito1574
    @litevenito1574 2 ปีที่แล้ว

    Your video helped me for understanding the concepts of color space. Thanks!

  • @luanaferreira9468
    @luanaferreira9468 2 ปีที่แล้ว

    Excelente! poderia ter um link para o código no github :)

  • @swagatapaul5351
    @swagatapaul5351 2 ปีที่แล้ว

    Wish you showed a 2-D implementation

  • @客家饒舌執牛耳
    @客家饒舌執牛耳 2 ปีที่แล้ว

    thankyou , well explained!

  • @abhishektyagi9091
    @abhishektyagi9091 2 ปีที่แล้ว

    very good

  • @emreparlak2729
    @emreparlak2729 2 ปีที่แล้ว

    great video. i wonder if it is possible to make the fft function using regular for loops instead of recursive function version

    • @aahaanchawla5393
      @aahaanchawla5393 ปีที่แล้ว

      I reckon you could, u can view it as a binary tree which has the odd incices and even indices as it's two children and then use bfs or dfs to traverse everything, without recursion

  • @anbuingoc4495
    @anbuingoc4495 2 ปีที่แล้ว

    Could you please give me a jupyter source code?

  • @aminramezani9947
    @aminramezani9947 2 ปีที่แล้ว

    Great explanation. Thanks!

  • @diegocaixeta8848
    @diegocaixeta8848 2 ปีที่แล้ว

    Boa tarde Moacir. Hoje recebi da UFV um convite muito especial. Falar sobre a minha graduação. Será publicado lá na revista na parte do ex-aluno. Não pode esquecer de você e de como você foi importante na minha graduação. Nem gosto muito de Python, mas poxa vida... assistir uma aula sua novamente da uma nostalgia impar. Abraço professor! Aqui.. minha esposa está pra ser chamada em um concurso da UFV. Se nós formos pra lá vou poder voltar aos meus estudos. Me chama pro mestrado? kkkkkkkkkkkk Fica com Deus professor! Tudo de bom pra você hoje e sempre!

  • @vkr9467
    @vkr9467 3 ปีที่แล้ว

    This is excellent stuff on fft, very clearly explained. Can we access this jupyter notebook? Thanks and best regards

  • @pedroernesto7821
    @pedroernesto7821 3 ปีที่แล้ว

    Muito bom o vídeo!