Deep Learning 54: CNN_6 - Implementation of CNN from Scratch in Python

Показать описание

This lecture implements the Convolutional Neural Network (CNN) from scratch using Python.

#deeplearning#cnn#tensorflow

Рекомендации по теме

Комментарии

This is the best tutorial i have ever seen... Thank u so much

eromoseleodanlumen

You have done very well. I learn a lot. However, in Conv_op backprop, why you do not calculate dL_dx so we can backprop it to previous layer? and no bias. It seems that input of 1 channel is easier to calculate. If it is the color image, should we sum the 3 results of conv from R, G, B to 1 value in each new channels of the output?
In softmax layer, I read somewhere that if we reduce the image size, then we should increase the channels (filters) so no data lost. And I read somewhere that dL_dz can be calculate directly by x (1-x) to accelerate.
I hope to see your video about Batch Normalization forward and backward. This layer increases training result and reduce time a lot?

Thank you very much for your video.

nhactrutinh

Sir great work the way you teach math behind working of neural networks is incredible 👏🙏

malaypatel

Sir, thanks for the video. I have further modified your code to enhance the accuracy

johnparker

Why the return statement is inside the for loop in backward propagation in Softmax Class ?

ankurchourasia

you r very nice teacher. Thanks a lot. If possible please make videos on transformers, RNN and LSTM

rahulrai

in the calculation on convolution filter, I think we might need another loop to loop over number of channels, probably needed to add on forward_prop function

ravimohanreddydwarsala

Hats off, great content @Ahlad Kumar, thank you for everything you are doing. I'm ML master student and the videos helped me a lot in my DL course. You should get paid for your efforts, you should create a patreon campaign.

ndobrevgw

can you please share notebook of this code, i write the code but its shows error

dharmvirdharmacharya

Do we have to always convert RGB images to Grayscale or u just did for easy computations?

rajeshwarsehdev

shouldnt be the loss at 24:27 => - np.sum( label * np.log(out_p) ) ?

comandernehal

can u plz help with the code for "speaker recognition using CNN"

priyankaabhang

awesome tutorial. please do you have a tutorial on how to implement transfer learning using inception v3?

epaphraspeter

Informative videos. Please carry on Sir.. It's really help us and we need more videos

bibhunanda

could you also just zero pad and loop through all pixels instead of the generator function?

andrewkreisher

I can't thank you enough for this content. I was looking for a demonstration of creating a CNN from scratch without the aid of a neural network framework. This seems to be it. I only have one question so far, that is, where can I download the code. I looked at the github site under your name and could not find it. Where can I find either a python or python notebook version of this code?

theminertom

when i run this code i downgrade my tensor to tensor 1.2 but still face this problem(InvalidArgumentError (see above for traceback): You must feed a value for placeholder tensor 'Placeholder_2' with dtype float
[[Node: Placeholder_2 = Placeholder[dtype=DT_FLOAT, shape=<unknown>,

oriabnu

You are a god sir. Thank you so much. Keep it up :))

neillunavat

Sir, This implementation is not consist activation function.please make a lecture with activation function.

kpgpt

Hello Sir, thank you for the hard work. This video was very helpful to me

vinitsingh

Deep Learning 54: CNN_6 - Implementation of CNN from Scratch in Python

Convolutional Neural Network from Scratch | Mathematics & Python Code

MIT 6.S191: Recurrent Neural Networks, Transformers, and Attention

MIT 6.S191 (2023): Convolutional Neural Networks

MIT Introduction to Deep Learning | 6.S191

Intro to Convolution Neural Networks(CNN)|Deep Learning #6| Tensorflow and Keras|

Build a CNN in 6 Minutes

Simple explanation of convolutional neural network | Deep Learning Tutorial 23 (Tensorflow & Pyt...

Padding in Convolutional Neural Network

Deep Understanding of MNIST problem and its CNN solution using CNN Explainer

Hot Dog or Not Hot Dog – Convolutional Neural Network Course for Beginners

Lesson 6: Practical Deep Learning for Coders 2022

Deep Learning L04 : Neural Networks: Convolution, Pooling and Visualization

🔴Deep Learning Free Full Course 6 Hours

Using Deep Learning to Classify Musical Instruments

Deep Learning Module 3 Part 6: One Layer CNN

MIT Introduction to Deep Learning (2023) | 6.S191

MIT 6.S191 (2018): Convolutional Neural Networks

GANs, CNN Architectures, Deep Learning on Graphs and Henry AI Labs with Connor Shorten || TAIWS #6

CNN and RNN in Deep Learning | Deep Learning Tutorial | Deep Learning Algorithms | Simplilearn

Deep Learning Full Course - Learn Deep Learning in 6 Hours | Deep Learning Tutorial | Edureka

Convolutional Neural Networks: Unlocking the Secrets of Deep Learning

The Essential Main Ideas of Neural Networks

MIT 6.S191 (2021): Convolutional Neural Networks

Neural Networks and Deep Learning Complete Course