Dogs vs Cats Image Classification (CNN) | Deep Learning | Python

[@HackersRealm]

Hey Hackers, I have updated the code to test the model with real image. You can find the code in my website or GitHub. Happy Learning!!!

[@lebaminhdat2402]

Thank you so much for the video, without it I wouldn't be able to pass my final project, wish you can develop your channel further in the future

[@kwanshinoda7859]

very true bro

[@HackersRealm]

Glad it was helpful!!! I will try my best to create more content!!!

[@lebaminhdat2402]

@@HackersRealm many thanks you again

[@PraveenKumar-ob7gd]

@@lebaminhdat2402 u do that project u create ppt slides is available to send me plsss..... I want 😊

[@lebaminhdat2402]

@@PraveenKumar-ob7gd sorry for reply too late, can you give me your email or something I can send to you!!

[@Tis_I98]

Awesome class Very clear abd understanding

[@myhaluu4277]

Thanks for your video clip, it helped me a lot of in my graduation thesis

[@HackersRealm]

Glad it was helpful!!!

[@vikrammane-xu6qn]

What kind of pre-requisites do i need before going through the project? except python basic syntax

[@HackersRealm]

You can watch it as it's basic deep learning project

[@kirushanthuthayakumar7769]

Its so nice and very helpful. Thank you so much ❤

[@HackersRealm]

Glad it was helpful!!!

[@erwanerwan6196]

hello sir !!! i am still on it...i tried to do the same code on jupyter nb CREATE DATAFRAME ETC...But i find 1 for any sample cats dogs why..?? do you have any idea to fix this issue ?

[@ahmadshabaz2724]

awesome tutorial bro. lot of respect

[@HackersRealm]

Thanks for your kind words!!!

[@fatmanursefer1482]

Can I do the same operations for more than 2 classifications?

[@HackersRealm]

No, there should be some changes done on the model side for that but you can still try

[@jimgolias9745]

Hello and thank you for the video! Why from keras.preprocessing.image import load_img gives me an error? How do i fix it?

[@HackersRealm]

I think, the module may get updated in some version, try to check how to import the module for that version, it will fix the issue.

[@sadikaljarif9635]

is it possible to use this for alexnet model???

[@HackersRealm]

yes, you can use any model for this

[@sadikaljarif9635]

@@HackersRealm what about image procesing ..if i use alexnet i have to change it right???

[@HackersRealm]

@@sadikaljarif9635 a little, may be, it depends on the input size you're specifying, or you can use the same steps for alexnet

[@sadikaljarif9635]

@@HackersRealm ok thanks

[@jimgolias9745]

Hello and great job on the video. I get an error in plt.figure(figsize=(25,25)) that says KeyError: 'images' . How do i fix this?

[@HackersRealm]

the column name is not present. Are you using the same dataset from the video?

[@maedeha.b4714]

I am having the same issue

[@ashmikajayharinkhede5197]

10k subs incoming!!!

[@HackersRealm]

Hope so!!!

[@erwanerwan6196]

ok with class_name it's better ..no invalid syntaxe now !! ..because i ve seen you added it ...

[@Kaijoh]

haarcascade ? to really differ both of them using webcam

[@HackersRealm]

didn't get you, can you rephrase the question?

[@saitarun6562]

Thanks Liked & Subscribed ❤

[@HackersRealm]

Thanks mate😄

[@amaobirufusbobo3707]

Your explanation was superb, but i noticed there is no prediction in this project and how can I add it if possible

[@HackersRealm]

You can see other similar projects in the playlist, those videos will have prediction part

[@sumaiyasafdar6845]

i wanted to ask that how can we present this project in our college? like do we need to write frond end too? of just the code in google collab??

[@HackersRealm]

you can show the result by giving the image input to the model and display the result

[@ghitalazrek4759]

i have this ValueError: Input 0 of layer "max_pooling2d_8" is incompatible with the layer: expected ndim=4, found ndim=5. Full shape received: (None, 2000, 78, 78, 32)

[@HackersRealm]

I think there is some problem with the reshape of input. You can see the None at the start. Try to reshape it and check whether you have the shape (2000, 78, 78, 32)

[@kaziranga_national_park]

Hello sir, apporox images 18 lakhs, how to Wild animal images classify and automatically segregation move on folder wise. Please help

[@HackersRealm]

if you have class labels like this dataset, with few changes you can do it!!!

[@kaziranga_national_park]

@@HackersRealm Sir, i am begginger, please video provide. I am from Kaziranga National park (Assam) India. data oparator.

[@kaziranga_national_park]

@@HackersRealm no idea.

[@deeber35]

I'm getting an invalid syntax error on this line {which matches your code}:: "for class in os.listdir("dog-vs-cat-classification"): ^^^^^ SyntaxError: invalid syntax The next line in the code gives same syntax problem . Any ideas why? Thanks

[@kaziranga_national_park]

Are Spyder software is working. this Coding

[@HackersRealm]

you can use this code in spyder as well

[@priyabratmohanty6610]

For the generators I am getting "found 0 validated images file names belonging to 0 class" and same response for class 1. Could you please help me with this, why I am getting this . plz help, I have my project submission in 2days

[@HackersRealm]

Are you using the same dataset and in colab?

[@priyabratmohanty6610]

@@HackersRealm yes the data set is same only for my project. However I am using Kaggle notebook bcz my laptop is a bit slow. So is it happening because of this !?

[@HackersRealm]

@@priyabratmohanty6610 yes you could you colab or you can update the folder path and use it in kaggle

[@merosinchan8048]

Hello. Can I ask why my my dogs and cats images can load, at first and afterwards it’s just figsize x figsize = so,e digits output? And my accuracy is not moving :[

[@HackersRealm]

you're saying the images are not loading?

[@rushikeshchavanke9198]

29:19 testing time stamp

[@darthyrabecka1472]

Why length of values doesn't match length of index

[@darthyrabecka1472]

plz reply

[@HackersRealm]

@@darthyrabecka1472 is it showing error while running the code?

[@HackersRealm]

@@darthyrabecka1472 it's usually have same values only

[@tanvivarak6185]

Hello i really need your help. After successfully running this code I ma trying the prediction part which apparently I am not able to. i tried every possible way that I could. plz also provide the prediction code.....i really need it

[@HackersRealm]

you have to load and resize the image based on the training we have set. img = np.array(PIL.Image.open(image_name)) pred= model.predict(img) If you face any issues, reach out to me in linkedin or insta, links are in the description

[@jervx829]

img = cv2.imread('imagehere.jpg') img = cv2. resize(img, (128, 128)) /255. res = model.predict(np.array(img).reshape(-1,128,128,3)) print(res)

[@erwanerwan6196]

hello i run only the first loop and get invalid syntax i triple checked it but still same error how possible ?????

[@HackersRealm]

which loop you're referring to? Could you give more detail on that.

[@erwanerwan6196]

@@HackersRealm glad to hear from you !i thi s is my code !input_path = "parentfx" label =['cats','dogs'] for class in os.listdir('parentfx'): for path in os.listdir("parentfx/"+class): if class == "Cat": label.append(0) else label.append(1) input_path.append(os.path.join("parentfx ", class , path )) print( input_path [0] ,label[0] ) File "", line 4 for class in os.listdir('parentfx'): ^ SyntaxError: invalid syntax

[@erwanerwan6196]

maybe TF version is the issue ?

[@fahadhos]

@@HackersRealm i have ques is it the actual binary classification , i just curious i thought i have to put the dog image then model will tell that this is dog

[@HackersRealm]

@@fahadhos yes it's a binary classification problem as the output is 2 only

[@cuocsongquanhta5398]

Please export model.h5😢

[@HackersRealm]

You could train the model quickly using colab itself

[@deeber35]

What do you do if you have hundreds of corrupt files, like in the kagglecatsanddogs_5340 files?

[@HackersRealm]

You could automate the cleaning process in a loop if the file is not opening

[@deeber35]

@@HackersRealm Which opening of an image command would you advise I use to find the corrupt ones?

[@HackersRealm]

@@deeber35 any library will throw error if you try to open a corrupted file

[@erwanerwan6196]

HI I have only 48 images not more.....batch size =20 then model .fit does not work Epoch 1/50 why ?? --------------------------------------------------------------------------- InvalidArgumentError Traceback (most recent call last) in 4 epochs =50, 5 validation_data =validation_generator , ----> 6 validation_steps = 1.2,

[@HackersRealm]

what is that validation steps parameters?

[@aditya_r_k__]

When I'm trying to predict some results by this model it's going to show [[0.]] which makes no sence

[@HackersRealm]

is it showing the same for all the images?

[@aditya_r_k__]

@@HackersRealm usually the output will be [cat percentage, dog percentage], but in this case the output will be like [[0.]] that I tried 2 images showing same result

[@HackersRealm]

@@aditya_r_k__ Are you getting the proper validation score result; If yes, most probably the error may be in the loading of test data

[@tanvivarak6185]

plz provide the prediction code

[@phanthong0212]

Dataset link : 404 Not Found. Please fix it !

[@HackersRealm]

I will check it

[@HackersRealm]

www.microsoft.com/en-us/download/confirmation.aspx?id=54765 try downloading the data from this link

[@phanthong0212]

@@HackersRealm Thank you so much !

[@foggypoggy6143]

there is nothing called load_img in keras.preprocessing.image

[@HackersRealm]

the modules might be updated.. you could use keras.utils instead. keras.io/api/data_loading/image/