For those that are frustrated by the first line of code not working, just remove the "tensorflow-gpu" part and run the code. You may want to skip the GPU setup part cause it won't work anyway. The algorithm will work, except slower (mine is 656 ms which takes me less than 1 minute to go through the entire epoch of 20)
@ricefield350
7 күн бұрын
hey, do i only remove tensorflow-gpu or the tensorflow too? because tensorflow is not working on jupyter notebook
@Jey187
7 күн бұрын
@@ricefield350 I think tensorflow should be working. You only have to remove the GPU one
@KarrsonHeumann
Жыл бұрын
I really love these longer tutorials. You explained things so well in this one that I feel like AI development finally clicked for me, not just in terms of this specific application, but also in general. I would understand if you'd be worried about length vs entertainment, but honestly you teach so well and you are so enthusiastic I don't think that should even be a concern. Thank you so much! :)
@lemon_and_berry
5 ай бұрын
Agreed !! Waiting for such vids
@atharvmunot8305
5 ай бұрын
Thank you so much for this Tutorial!! IT IS THE BEST !! P.S. A side note for the recent viewer, while compiling the model, use the command: model.compile('adam', loss = tf.losses.sparse_categorical_crossentropy, metrics = ['accuracy']) This change caters to the recent change in the naming conventions and ensures that the saved .h5 model runs when loaded
@donguyengiac5046
5 ай бұрын
OMG you're a lifesaver!
@Sneha-e7d
2 ай бұрын
@atharvmunot8305 hey!! Can you help me with installation.. actually I am having problem while installing the tensorflow gpu , its saying python version is not compatible , do you know any other way of installing that without degrading the python version.
@dimasalangxt3482
2 жыл бұрын
Amazing job on these videos! Would love to see a tutorial featuring 9 or more classes, thanks!
@leonp5281
2 жыл бұрын
I second that!!
@qualitycontrol716
Жыл бұрын
please Nicholas
@danielleivy8180
Жыл бұрын
??? I just kept adding classes, when it hits 9 it just moves onto 10....
@poisonedexistence3027
Жыл бұрын
Yes, Nicholas please! multiclass tutorial from you is needed=) Thank you
@salvinprasad8592
Жыл бұрын
Absolutely brilliant. I will use this structural approach in my third paper for my PhD. Thanks so much
Жыл бұрын
Hello Nick, thank you for this awesome tutorial, I learned a lot. I was wondering if you published another tutorial with more classes involved? (at 13:01) Thanks
@FandangoJepZ
2 жыл бұрын
would really appreciate one with more classes! Trying to make an AI for SET
@mohamedgaal5340
6 ай бұрын
Thanks a lot Nick! I like how you skim through the mathematical concepts behind your code. Very informative! I'm watching the whole playlist :)
@nachoeigu
2 жыл бұрын
Another way to extract images from the web is by using web scraping and it is more robust and flexible technique.
@Christian-dd2qm
Жыл бұрын
Great content and I love that you speak proper English! I am not a native speaker and had my fill of Australian and Indian accents.
@bl4z3_kanazaki
Жыл бұрын
Dude makes DL actually fun to learn! I can't learn anything from the lecturers at my college because they talked to much and didn't even explain anything! Thanks man
@mahendrakergaurav5867
2 жыл бұрын
Amazing Tutorial, highly underrated channel, will share this with my friends.
@aminaleali7161
10 ай бұрын
Exceptional talent for teaching! Informative, clear, and I love the pace of it. No fluff and to the point. Thank you and great job!
@BHPFestival
6 ай бұрын
Hey mate, we need an updated video on how to install gpu capability for jupyter notebooks in visual studio code. Tensorflow has updated and some in this tutorial no longer works. Without gpu usage I don't need to tell you ai is out or range. I've seen so many other videos that get so messy you can't be sure where you are. I think you can be the one to clear it all up is a simple and beautiful way. Otherwise - your video is wonderful.
@housemdaaji4884
2 жыл бұрын
Well explained!! Would love to see you do the same for satellite imageries (crop identification, urban change detection,etc)
@alexiadeboynes4656
Жыл бұрын
yes please!
@lubwamaemmanuel7743
6 ай бұрын
Hey Nicholas, that is an amzing tutorial, i really learnt tonnes to take me to my next learning of ML. Thanks so much.💯
@brunospfc8511
2 жыл бұрын
Thanks Nicholas, i'll try it
@ocraking
4 ай бұрын
amazing video and amazing teacher. We would love if you can explain callbacks in another video, thanksss
@emiliani8side
11 ай бұрын
Superb and well detailed video! It would be amazing to see you breakdown image classification through multi-classification rather than binary with maybe 4 different datasets? Also, a confusion matrix to display values at the end would also be extremely helpful.
@photorealm
Жыл бұрын
Awesome video. Love the way you explained all of the steps in great common sense detail. 5 Stars 😊
@ziddimalikziddimalik1288
Жыл бұрын
Dear sir, Your video is so awesome and you deliver each point very clearly and it need more video related this topics and student want to be more learn to your channel I hope you will be share more video such kind of work... Good job sir👍
@yusufkemaldemir9393
Жыл бұрын
Who are you bro:))) This is best training video I have seen on KZitem.
@numforfrank9205
2 жыл бұрын
Your videos are top notch, explicit and yet humorous at the same time😅. YOu make learning AI easy. Thanks Nic.
@ahmedbahpullo8544
2 жыл бұрын
C'est ce que je voulais savoir Thank you you are the best, 💪💪💪
@NicholasRenotte
2 жыл бұрын
Thanks a mil @Ahmed!
@filixzs
Жыл бұрын
Amazing explanation, Im using this for thesis project, I'll let you know how well it went 👍
@ameer-alahmadi
Жыл бұрын
The great explanation I've ever seen! Thanks a lot!
@dimitrimattheis3740
Жыл бұрын
awesome tutorial! would love a video on how to use more than two classes
@josepalacios843
Жыл бұрын
The thumbnail says "This is proably terrible!" but the content hits like boulder. Nice work man. Thanks for sharing your experience.
@thecodegobbler2179
Ай бұрын
Thank you for working through your mistakes!
@kiwppp6347
2 жыл бұрын
Everything works, especially free, use until it is too late, guys !
@ajiteshshree9914
2 жыл бұрын
Please make a video on how precision, recall and accuracy work.
@ll-shellshock-ll1544
Ай бұрын
Hi Nicholas, I have an issue with saving my pretrained EfficientNet model. Essentially, when I saved the model and imported to test it, I got a completely different result from the original model. Theoretically, the saved model should have the same, if not, similar accuracy to the original model. I tried to debunk this problem but could not figure it out. Does anyone here encounter this problem? If so, please let me know how to resolve this. I thought this would be a great content to cover for you, Nicholas. Keep it up with the good work. You've helped me a lot with my portfolio projects. Thanks
@mikohalurangersid-green493
2 жыл бұрын
Amazing video with perfect explanation I wonder if you can make some kind of tutorial with image classification using RNN in the future
@amirhosseinkarimi1760
10 ай бұрын
this is very well done sir! thanks for the great content!
@_leonard_6173
Ай бұрын
45:50 - I am confused. What do you mean by "shuffle" ? I have the data just like yours : Data folder containing a folder for each class. Also another question that I have is : if my Data folder contains class A (50% of parent folder), class B (50% of parent folder), and I make the training set 70% of the data folder just like your code does, will it be 70% of each child folder or 70% of the entire Data folder?
@khenpahilanga9596
Жыл бұрын
Nicholas, THAAAANK youuu!!
@donystark5
Жыл бұрын
Tu es le meilleur broth 🤝👌👍
@natgenesis5038
6 ай бұрын
Thanks so much for your help of my project
@sasu_chann9188
Жыл бұрын
tysm for the video it's so helpful, I'm wondering if you could make a video about the difference between image classification and image retrieval and how the code differentiates (what to add or remove).
@agustinsaavedra4273
11 ай бұрын
Thanks from the example.
@ersinbicak7448
Жыл бұрын
In 42. row when I run image_dataset_from_directory('data') code. It gives unicode error. It says 'utf-8' codec can't decode byte 0xfc in position 32: invalid start byte" I can not forward for this issue. Can u help ?
@takomensei3710
9 ай бұрын
im having the same problem, did you find the fix?
@RC-dx6vv
4 ай бұрын
me too, unfortunately... still scrolling through comments hoping somebody posted a solution.
@Onouphrius
2 жыл бұрын
Great video! So, what i was wondering is what all would i need to change for a scenario where i have more than two classes, 27 in my case.
@ادهمالصلاحي-ف6و
Жыл бұрын
Hi bro My case is similar to your case Did you solve this problem ? I need to classify 36 class
@Musa-bt8wt
Жыл бұрын
@@ادهمالصلاحي-ف6و you don't need to change much, just create extra directories with the image data in them
@ادهمالصلاحي-ف6و
Жыл бұрын
@@Musa-bt8wt I got problem in evaluation part I can't run the metrics precision, Recall , Accuracy I got an error
@shayp9428
Жыл бұрын
@@ادهمالصلاحي-ف6و found a solution?
@Nnemka
10 ай бұрын
Thank you! You are the best!
@vishalvaibhav18
Жыл бұрын
Great video🎉. Keep it up. Please.
@parthgupta9338
Жыл бұрын
Thank you so much. I learned a lot. 😀
@IksanOktavRisandy
4 ай бұрын
58:20 I want to ask, what if i have 19 classification , should it be Dense(18, activation="sigmoid") so in sigmoid the range from 0 to 18 (19 values) ?? wait i just found that sigmoid activation ONLY have 2 outputs that 0 and 1, so should i use "softmax" activation instead?
@alirezanorouzi8924
Жыл бұрын
thanks for this . i realy enjoy it
@aditiarora2128
Жыл бұрын
sir plz share image classification using online image augmentations well
@PJ-nc4jh
Ай бұрын
So im using a M2 chip Macbook pro where the CPU and GPU are integrated. Whenever I run the length of GPUs it states 0 and then the set_memory_growth() function is useless. When I check for CPU, it states 1, so will that still create an OOM error in the future?
@georgewang7770
Жыл бұрын
Thanks for the tutorial. Quick question: I see you took steps to preprocess the images. That's great. But what if we took each image as simply pure raw data (not even treating the files as image files), and just run CNN on them as is. Would the classifier still work? Maybe not as accurate, but it should still work right? From CNN point of view, it will just detect raw data pattern as is, and still reach that final single node as 1 or 0.
@muhammaduseram9405
7 ай бұрын
very well explained , one question can we do the same process for commonly lost items, context is we are building image matching loss and found system where there is need of image classification to generate tag for each image and put them into its relative category, can we use same network setting, waiting for you reply ASAP, Thanks
@sarveshkanwar1508
2 жыл бұрын
thanks bruv your video was really helpful for my HACKATHON
@sudhakarm4573
Жыл бұрын
hi.. nicholas. very clear explanation.. Thank you so much... I have a doubt. whether this model could be used for our data, if the accuracy is good. rather than pre trained models... In other words, if we can give good accuracies just by sequential arranging of layers, what is the neccessity of pretrained models or transfer learning. please clarify this .. And I want to use more than 2 classes. How could be do so?. or Anyone please clarify
@eatonasher3398
Жыл бұрын
@46:00 Where is the "shuffle" function called? Wouldn't it shuffle each time you take a partition? If so, what good does skipping the first partitions do?
@50CenT123smoke
Жыл бұрын
Hi Nicolas. Some truly amazing work. I learned from this video what I couldn't learn through 8 lectures of 3 hours each :) Is there a possibility to add a confusion matrix to this as well
@priyam66
Жыл бұрын
yes , you can add a confusion matrix as well import tensorflow as tf tf.math.confusion_matrix
@shakilahmad2640
Жыл бұрын
hi, did you add a confusin matrix ?
@evelynong3483
Жыл бұрын
do you have a tutorial on a similar classifier but with multiple (3 or more) classes? for example, maybe a happy/sad/angry face?
@LuffyMonkey0327
Жыл бұрын
please do a tutorial of this with google collab. trying to do this in google collab. am having trouble
@nostalgiccringeallhailchel3881
Жыл бұрын
44:00 watch again
@massiliabou345
6 ай бұрын
Mach'allah thank you very much 🌻
@shattikbandyopadhyaa1787
8 ай бұрын
can anyone give me the link to the video (Setting up stuffs in the beginning) he mentioned about?
@ragirikarthik2057
7 ай бұрын
Do you know how to setup
@RashmiBajaj-mf3el
9 ай бұрын
Hi, I implemented the code just as you. But at the end, it is not predicting happy and sad correctly. You only checked for one image. Even your code wouldn't work ig, I do not know what is the error, but you should correct it as well! Please le me know if you resolve it and how.
@SoulFrmTitanic
6 ай бұрын
No, it is completely correct and it worked for me!
@Aditya-ms1ll
Ай бұрын
more epochs would work for you , or try changing optimizer and find suitable one asper you
@arnabghosh5547
2 жыл бұрын
Nich, would please also make theory explaining CNN, object detection, their metrics & hyperparameter tuing
@NicholasRenotte
2 жыл бұрын
Ohhhh man, theory isn't really my fav but I'll see what I can do!
@m.rishab4770
3 ай бұрын
@@NicholasRenotte can You please make video on how to do it for multiple classes ?
@mattzucca4102
Жыл бұрын
So, I got this to work. tensorflow doesn't need import tensorflow-gpu in case anyone goes down that rabbit hole.
@ramonatilanovillalobos6622
Жыл бұрын
Thanks you save my life
@mattzucca4102
Жыл бұрын
@@ramonatilanovillalobos6622 np, that was a rough one
@afkhoso
9 ай бұрын
thanks! had the same issue! comments and conversations like this help others very much!
@AlvaroPerezLuque
Жыл бұрын
Hi Nicholas, the tutorial is fantastic. There is a small bug in the code. The data object is always shuffling the data so there is no difference between training and validation data. To fix this bug it is necessary to change these lines: 1. Edit the line data = tf.keras.utils.image_dataset_from_directory('data'). It should be: data = tf.keras.utils.image_dataset_from_directory('data', shuffle=False,) 2. After this line add the following: data = data.shuffle(1000, seed=100, reshuffle_each_iteration=False) Without this correction the val_accuracy will always be 1 as there is the same data. (The problem arises when calling the take method because the data is reshuffled.)
@jamesmccannon7769
Жыл бұрын
Are you sure that this is correct? When testing this although this is stopping the accuracy becoming 100%, it's validation accuracey is 1.00 from the start??
@j220493
Жыл бұрын
There is another bug even bigger than that, he is doing data leakage. When he download the images from google, there are a lot of them duplicated or triplicated, so when he splits the data, same image is in all three divisions (train, val and test), for sure, so is its requeried to purge the data before all the process.
@leonp5281
2 жыл бұрын
I would love to use more than 2 classes as you mentioned on 12:55. What do we need to change on the code ? Any possibility for an updated .ipynb file if you don't want to create another tutorial? Thank you so much for the great guides!!
@pranavtiwari_yt
Жыл бұрын
have you figured out? plz tell if yes
@marekkamm1524
11 ай бұрын
pls tell me too :) @@pranavtiwari_yt
@RoccoEduxD2
22 күн бұрын
we need to see how to work with ore than 2 classes.
@binalweerasena753
10 ай бұрын
Now you don't need to pip install tensorflow-gpu. GPU features are already inside tensorflow main library. Trying to install tensorflow-gpu got me an error which ate up my time. Thanks Renotte for this amazing videos. You truly inspire us, the newbies.goodvibes
@afkhoso
9 ай бұрын
thanks! had the same issue! comments and conversations like this help others very much!
@TadiwanasheNyamapfene
2 ай бұрын
Thanks for commenting. I ran into the same error
@montaguemonro565
7 ай бұрын
I'm honestly confused on the stacking of the layers, what influences how the hyper-params and what influences how many times you're duplicating the same 2 layers?
@TheCertifiedman
5 ай бұрын
Hi I want to create a Model that learns where to place ground control points given images showing how to place them. The input is a kml and I want an output of coordinates
@rmarti5163
5 ай бұрын
@1:06:00 I guess I still don't understand how the training is working. How does it know if it is correct or not? what is the data loss? is it the data that doesn't actually show a happy or sad human face?
@venomlovekitties
Жыл бұрын
As a non coder person I instantly subscribed because of the simplicity you showed by your teaching skills. Thanks man, love to see more content from you.
@CantPickTheNameIwant
Жыл бұрын
as a CV engineer, I instantly hit the dislike button under this video
@mihai3678
Жыл бұрын
@@CantPickTheNameIwant that’s what I wanted to say 😂😂big source of misinformations on this channel, specifically in this video
@daryladhityahenry
Жыл бұрын
@@mihai3678 Can you tell which one is misinformation and how should it be? So I can know which one that I should look for... THank you....
@akikabanu5885
10 ай бұрын
@@mihai3678 how come? do you think you could explain?
@muhammaduseram9405
7 ай бұрын
@@CantPickTheNameIwant at least you should clear your point if you said it
@Faridas-wy4kx
5 ай бұрын
Can i ask one question? Can i change the rgb format color to CIELAB and use it for CNN?
@chetankumar5634
2 жыл бұрын
@Nicholas Renotte Can you make a video more more than 2 classifier?
@ankitapal9229
2 ай бұрын
very useful and explanation is also just awesome..Thanks for the tutorial..
@marti-nz
2 жыл бұрын
This tutorial is amazing, not only are instructions easy to follow but sufficient explanation is provided so I know why each line of code was added. Great Job!
@DinaDina-xf8nx
6 ай бұрын
Thank you so much ❤.l hope to give us ather deep learning project of multiclassfication image problem
@SummerSausage1
Жыл бұрын
Now that the model is created, how do we actually update it in the future? Like training the model pt 2 with even more data? any video on that?
@TheSakyoGamer
2 жыл бұрын
This. Was. AMAZING! Oh my gosh. Thank you for such for this tutorial. I've been wanting to get into machine learning for so long, but never knew where to start or how to work these models. With how long this video was and how excellent your commentary was, it helped so much! I plan to watch a ton of your videos about creating some more models.
@meghanah7640
Жыл бұрын
Sir If there are more than two class then how to build a deep learning model..can u guide me please.. it's so urgent 🙏
@bhairafeek
Жыл бұрын
Could you please provide the video link for multi class images(More than 2 classes) @NicholasRenotte
@aviparnabiswas3707
2 жыл бұрын
Hello Nick, was wondering can CNN be used to classify the before and after measurement of lawn grass cutting? I was wondering how well this model will work with such images.
@niharpatel8520
2 жыл бұрын
Hello I’m having an error as UnimplementedError: Graph execution error How can I resolve it,please guide me
@ashfeenaziz3330
2 жыл бұрын
why I can't install those frameworks, It end with saying doesn't satisfied the version .
@thepresistence5935
2 жыл бұрын
Hey, Bro. Please make a video about Zero short learning for Images. Please
@AstroDeeptej
6 ай бұрын
for the deleting images less than 10KB 16:37 use this script instead. import os, os.path for root, _, files in os.walk(os.path.join(data_dir, 'sad')): # replace sad with happy for happy for f in files: fullpath = os.path.join(root, f) if os.path.getsize(fullpath) < 10 * 1024: # set file size in kb print(fullpath) os.remove(fullpath)
@rbrowne4255
2 жыл бұрын
Thanks great tutorial!!! are there any advantages of gray-scaling the images as part of the pipeline?
@NicholasRenotte
2 жыл бұрын
Reduces the number of parameters, it's very much dependent on the dataset! If color is likely an important determinant of a class then I probably wouldn't grayscale!
@lancemarchetti8673
9 ай бұрын
Nice! Imagine if we could build a classifier that can spot Base64 in a screen capture and extract it accordingly. In digital Forensics this could be quite handy in cases where base encoding is used to hide particular image data.
@ahmedbahpullo8544
2 жыл бұрын
can you explain the classification of emotions with transfer learning using the efficientnet model for example. and an emotion database such as fer2013 or ck+.. and thank you very much for this wonderful video
@heatskill5806
2 жыл бұрын
Excellent video, it helped me a lot. Thank you very much. By the way, in the case of 3 outputs, the number of neurons in the final layer would have to be changed and instead of the sigmoid function, what function would it be?
@priyam66
Жыл бұрын
The activation function would be Softmax in the case of multi-class classification. For binary classification, the activation function is Sigmoid.
@vikashkumarbangar63
12 сағат бұрын
Nich, Why use tensorflow pipeline for just training and testing of so less pictures. Due to this your model is getting over fitted as the accuracy, precision and recall all are equals to one.. i think the model will ambiguous result for any outside data.
@archniki_
Жыл бұрын
33:00 poker face when watching non face images (and actually happy one includes sad ones too)
@ivankuseta2174
2 жыл бұрын
In case you are still wondering why you received such a low max for the batch when preprocessing data, it is because you accidentally scaled the data twice. So when using the first approach of scaled_data you scaled them to max=1.0 and max=0.0, and then again when using the map function with x/255, you actually scaled it to 1/255 = 0.0039.... EDIT: Wow, I commented this right before you said you might have scaled it twice hahahahahah
@nikolamatijasevic9454
2 жыл бұрын
Can you explain in a little more details? Thanks
@danielagomez722
2 жыл бұрын
Hi! If I recreate this but have 12 classes would it work? I have no knowledge about image classification models so I'm sorry if it's a dumb question
@pravallikaa3306
Жыл бұрын
Yes it would, but honestly Idk what changes we'd have to make to the code
@yohanesyilma2797
5 ай бұрын
how to upload a folder from a desktop to Google Colaab?
@vool6042
4 ай бұрын
You have to convert it into zip file and then upload the zip file in google colab. And extract the zip file .
@Gregg69420
11 ай бұрын
Please show us how to put something like this into a GAN so we can make a promptable GAN.
@prasadk6510
2 жыл бұрын
This is excellent to classify any two classes of objects.But why we get error in multi-class however.Mismatch in input and output class.
@CHARLYFIVE7
Жыл бұрын
It's almost the same. You just need to change some parameters and the activation function in the output.
@djs4246
Жыл бұрын
Epoch 1/5 --------------------------------------------------------------------------- InvalidArgumentError Traceback (most recent call last) in () ----> 1 history = model.fit(train_generator, 2 batch_size=16, 3 epochs=5, 4 validation_data=validation_generator, 5 validation_batch_size=16 1 frames /usr/local/lib/python3.10/dist-packages/tensorflow/python/eager/execute.py in quick_execute(op_name, num_outputs, inputs, attrs, ctx, name) 50 try: 51 ctx.ensure_initialized() ---> 52 tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name, 53 inputs, attrs, num_outputs) 54 except core._NotOkStatusException as e: InvalidArgumentError: Graph execution error: Detected at node 'categorical_crossentropy/softmax_cross_entropy_with_logits' defined at (most recent call last): File "/usr/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/usr/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/usr/local/lib/python3.10/dist-packages/ipykernel_launcher.py", line 16, in app.launch_new_instance() File "/usr/local/lib/python3.10/dist-packages/traitlets/config/application.py", line 992, in launch_instance app.start() File "/usr/local/lib/python3.10/dist-packages/ipykernel/kernelapp.py", line 619, in start self.io_loop.start() File "/usr/local/lib/python3.10/dist-packages/tornado/platform/asyncio.py", line 215, in start self.asyncio_loop.run_forever() i can not correct this error bro can you help me please
@kaziranga_national_park
Жыл бұрын
Hello sir I am a beginner, wildlife Trapped camera approx 20 lakhs photographs, how to image classify and automatically segregation. Please help
@coding_with_chip
2 жыл бұрын
just a heads up, typo: 1:09:50 - PART 4: Evaluating Perofmrnace
@akritirani4600
7 ай бұрын
So if I am doing the classification for multiclass then I cannot use "logs" function, what should I do in such case
@chauvoluuhuong7485
8 күн бұрын
I would like to apply AI in visual inspection process ( PCB manufacturing). Assume that the image isn't completed. Is there any way to archive 100% accuracy ?
Пікірлер: 645