Learn how to use TensorFlow 2 and Python for computer vision in this complete course. The course shows you how to create two computer vision projects. The first involves an image classification model with a prepared dataset. The second is a more real-world problem where you will have to clean and prepare a dataset before using it.
π» Code:
βοΈ Nour Islam Mokhtari created this course. Connect with him here:
π Get Nour's free Machine Learning job-ready checklist:
βοΈ Course Contents βοΈ
β¨οΈ (0:00:00) Introduction
β¨οΈ (0:01:21) Course outline
β¨οΈ (0:05:11) Whoβs this course for
β¨οΈ (0:05:35) Why learn TensorFlow
β¨οΈ (0:06:25) We will be using an IDE and not notebooks
β¨οΈ (0:07:25) Visual Studio Code (how to download and install it)
β¨οΈ (0:10:50) Miniconda – how to install it
β¨οΈ (0:13:23) Miniconda – why we need it
β¨οΈ (0:17:24) How are we going to use conda virtual environments in VS Code?
β¨οΈ (0:21:20) Installing Tensorflow 2 (CPU version)
β¨οΈ (0:29:56) Installing Tensorflow 2 (GPU version)
β¨οΈ (0:43:34) What do we want to achieve?
β¨οΈ (0:45:26) Exploring MNIST dataset
β¨οΈ (1:05:54) Tensorflow layers
β¨οΈ (1:09:44) Building a neural network the sequential way
β¨οΈ (1:27:22) Compiling the model and fitting the data
β¨οΈ (2:00:52) Building a neural network the functional way
β¨οΈ (2:08:33) Building a neural network the Model Class way
β¨οΈ (2:14:31) Things we should add
β¨οΈ (2:18:29) Restructuring our code for better readability
β¨οΈ (2:23:11) First part summary
β¨οΈ (2:24:12) What we want to achieve
β¨οΈ (2:25:23) Downloading and exploring the dataset
β¨οΈ (2:34:20) Preparing train and validation sets
β¨οΈ (2:53:37) Preparing the test set
β¨οΈ (3:10:17) Building a neural network the functional way
β¨οΈ (3:22:12) Creating data generators
β¨οΈ (3:31:39) Instantiating the generators
β¨οΈ (3:35:37) Compiling the model and fitting the data
β¨οΈ (3:40:34) Adding callbacks
β¨οΈ (3:52:08) Evaluating the model
β¨οΈ (3:58:04) Potential improvements
β¨οΈ (4:08:49) Running prediction on single images
β¨οΈ (4:23:05) Second part summary
β¨οΈ (4:23:56) Where you can find me if you have questions
—
π Thanks to our Champion and Sponsor supporters:
πΎ Wong Voon jinq
πΎ hexploitation
πΎ Katia Moran
πΎ BlckPhantom
πΎ Nick Raker
πΎ Otis Morgan
πΎ DeezMaster
πΎ Treehouse
πΎ AppWrite
—
Learn to code for free and get a developer job:
Read hundreds of articles on programming:
And subscribe for new videos on technology every day:
Just what I need for my masters ππ
Man I sometimes I go through this channel and there is just so much to learn that I kind of feel overwhelmed a lil bit haha. Thank you for this course !
Exactly the same..
LoL, very true. Especially as an IT guy, this channel reminds me every time “you don’t know anything” π
So much to learn and so little time βΉοΈ
Really need a tensorflowjs course from basic to advanced.
I am about to start my graduation project and I swear this is the course that I needed, you canβt imagine how happy I am π.
Fantastic. I am losing count of how many times I have said thank you and I am running out of of superlatives. The courses on here are just incredible.
DM βοΈ
Thank you very much for posting my course! For those who asked about the ML checklist, I will be sending it very soon, I did not expect this number of emails tbh, you guys are awesome!
If you don’t get the checklist by tomorrow then please contact me on LinkedIn or Twitter, the links are in the description above.
Why my progress bars in the Terminal create newlines to notify each iteration in every epoch, meanwhile yours terminal only display 1 lines of progress bar run at a time (new iteration replacing the old one)? What must I do?
Thank you, you are a great person π
Is there a specific spot we should put the if __init == ‘main’ function
Thank you is a very small word to express my gratitude to you, free code camp.
19:35 if you have trouble with this step, it’s probably because the terminal is using powershell as a default and not cmd when you install vs code.
@Fake Where is the settings search bar for this?
@Fake Thank you
@Fake you are a life saver!!!!!!!!!!!!!!!!!!!! thank you so much
@Fake bro u really are a lifesaver man, i have been searching for this so long
@Fake Thank you I’ve been searching for this one literally saving my time!
A great video, literally saved me from a huge amount of time and money for learning tf
Hello thanks for the great course, one thing I noticed at the end was when we create test, val and train generators it sorts the directories as strings instead of integers, that’s the reason why you got 12 when you checked an image from the directory number 2 at the end. I actually couldn’t find the exact spot where we could solve that problem wisely and I was so lazy to come up with a decent solution tbh so did smth like this.
Created a new dict:
prediction_dict = {
0 :0,
1 :1 ,
2 :10,
3 :11,
4 :12,
5 :13,
6 :14,
7 :15,
8 :16,
9 :17,
10:18,
11:19,
12:2 ,
13:20,
14:21,
15:22,
16:23,
17:24,
18:25,
19:26,
20:27,
21:28,
22:29,
23:3 ,
24:30,
25:31,
26:32,
27:33,
28:34,
29:35,
30:36,
31:37,
32:38,
33:39,
34:4 ,
35:40,
36:41,
37:42,
38:5 ,
39:6 ,
40:7 ,
41:8 ,
42:9 ,
}
return this one from predict_with_model function instead:
predictions = prediction_dict[np.argmax(predictions)]
I hope that helps to the ones also experience the same thing
Good catch Eren, thank you! I will paste your comment to my pinned comment so that others can see it as well!
I can’t edit my pinned comment π
@Nour Islam MOKHTARI Thanks a lot for the course again Islam, no problem I guess the ones who face the same problem will end up looking for such comment eventually and therefore they’ll find my answer. π
great work. please upload more videos like this.
Since macOS doesn’t support tensorflow GPU, and for the miniconda available for macOS there is not script file in order to activate the miniconda on VS code terminal. I would appreciate if you can give some guild for those learning with macOS.
can you make a video on how to use javascript ,python and c++ and c and what software needed to use them and or how to get started as to using them for windows 10? Also can you provide insight on what can they be applied to, such as IP cameras, remote controlled devices, or anything that is a combination of electronics and coding.
I love IDEs. I hate notebooks. It doesn’t feel like code. Great videoπ
Beautiful course! Just completed it.
Thank you so much for such an exceptional tutorial on computer vision!!
Thanks for the training. Would love to see a condensed version of this with real world examples. I got a little lost in the code. Some more visuals would be helpful.
Extremely well done, presented and analyzed. Thank you so much!
One of the best tutorials in Computer Vision
Great crash course on TensorFlow, and even better instructor who’s able to articulate the essential concepts well. Thanks, Nour!