Emotion Detection using Convolutional Neural Networks and OpenCV | Keras | Realtime 

I was working on the same project for my end semester, I had gotten the code from elsewhere but couldnt understand anything, But your explanation is very clear. Looking forward to more such videos and thanks so much!

Very well explained, I should add just one thing, using keras tuner to tune the hyperparameters that are the number of filters and kernel size, would have given you better accuracy. Anyways, very well explained

Gotta love it. Play a noise for each detected emotion and see if you can focus on your work.

how many epochs should i train ? training stopped after 18 epochs . and how to generate model.h5 file ?

Hi, great video! I have to create a similar CNN for a school project but I'm required to use pytorch. Do you know any useful source for translating it?

Hi I'm getting error when I run Display Images File not found : [ Winerror 3_] The system cannot find the path specified , how to fix this error?

hey there i have followed the same steps in the video but my model is predicting only sad for any expression shown.. kindly help..

i am unable to save model, after retraining the model. Where is a problem? in checkpoint? pl guide.

FileNotFoundError: [WinError 3] The system cannot find the path specified: 'F:\\kaggale dataset\\imagestrain/disgust' sir i am using your code and got tthe above error again and again and i am working on jupyter notebook not on kaggle kernal...plz help me in using path

Hello, I faced a problem, I built a CNN model where the accuracy is almost 65%. I use that model in the pycharm code. But after running the code, when The webcam opens, it always shows the first label. The label doesn't change for a single moment. How can I solve this problem?

Thank you bro.Because of you I completed my AI project today

ive run it on mac..its only showing surprise emotion

Can you please guide us that which Tensorflow and Keras version are you using for training model. Great Explanation of the code!!

that's an explanation ever. thanks for ur sharing

getting error - cannot import name 'Adam' from 'keras.optimizers' (/opt/conda/lib/python3.7/site-packages/keras/optimizers.py) please guide

thank you so much it was great help!! I succesfully peformed emotion recognition.

Hi ,thanks for your vedio .its very help for my project. One doubt your code run successfully but model.h5 is not create why?

getting error while loading model... I think because of mismatch in the keras or tensorflow version between yours and mine so please tell me your tensorflow version i have tried to creat my own model but it is getting some problems

hey the training is taking so much time

Thank you very very much. this tutorial is the best.

I am gonna use this for making face detection project

where do i find my model to download?

I don't know if someone else has faced this problem. I trained the model but the model did not get saved. So, I made a little modification to the code. checkpoint = ModelCheckpoint("./model.h5", monitor='val_acc', verbose=1, save_best_only=True, mode='max') became......... checkpoint = ModelCheckpoint(filepath="./model.h5", monitor='val_accuracy', verbose=1, save_best_only=True, mode='max') Hope it helps.

Please reply to me immediately brother instead of using kaggle I used jupyter notebook to train images but I can't download h5 file from jupyter how do I download it from jupyter notebook please tell me

I am unable to get the model5 file in jupyter notebook how can i get it ?

Thank you so much for this awsome video!!! Helped me so much in understanding! Keep it up!!

My model is only predicting SAD emotion.. why??

ValueError: Input 0 of layer "sequential_2" is incompatible with the layer: expected shape=(None, 90, 3, 1), found shape=(None, 48, 48, 1)

thanks for the explanation and the file, it really helped me.

wich CNN architecture did you use in your project? VGG16?

great video. but when I run this code with my work it shows the error like this " ValueError: `validation_steps=None` is only valid for a generator based on the `keras.utils.Sequence` class. Please specify `validation_steps` or use the `keras.utils.Sequence` class." please help.

I have an error that I tensorflow/core/platform/cpu_feature_guard this tensorflow binary is optimised with oneapi deep neural network library to use the following cpu instructions in performance-critical operation AVX AV2X Please help me sir 🙏

I am in fifth grade this video is very helpful. i am working on this project and looking forward to more projects

I'm unable to get the model.h5 file

I have a doubt that weather we can use this same idea for detecting the crack in railway tracks? Can I train the CNN with the crack images and use this same process

What is the accuracy of the project?

This is my 1st ml project pls help what to show to when someone asked about the code to show where is the code? Can't i save this and use whenever I want?

Every thing was well explained... can you please make a sepreate video describing in detail about the modeling part with all the CNN and Deep learning concepts...

from keras preprocessing.image import img_to_array from keras preprocessing import image I am getting error invalid syntax error

How did you choose the layers of CNN , did u use some hyperparameter optimization?

Hi boss, please let me know, it is very long to run 48 epochs, maybe or then 16 hours and i m afraid that my cpu will burn motherboard and damage my macbook pro, so i stoped untill 4 epochs. How did you do? is it risky for my macbook to run the program many hours?

What is the final test accuracy(validation accuracy) you got???

I am getting an error exception has occurred: ModuleNotFoundError No module named 'tensorflow.python. Please can you help

Hi, Thank you for this helpful video. I also want to make the same project for my term project, but I face a minor issue at the 2nd last step "callbacks=callback_list". I am still trying to resolve this but I can't. plz, help me to resolve this issue. I am thankful to you for this.

I want to capture video in python sing OpenCV but not from webcam I want to capture video from live call during Zoom/whatsapp/Tewams or any other video calls tools to detect the face. can any help me, how i can do this. Thanks

Can we give the duration of each emotion in a certain period of time ?

can you tell me how to get model.h5 file because it is not saving in kaggle and whats the code for saving the model

showing an error while importing Keras optimizer what should i do?

I am not getting model.h5 file after following the steps in output. Can u help plz

hey, exactly what type of features are being extracted

Hi bro , In output data section there is no output file is produced why ? Help me plz

how much does it take to run epochs? In jupyter notebook for me out of 48 epochs each epoch is taking about 30 minutes

what are versions of packages/libraries you have used.?

hey, Very well explained and helpful video. while running the code, in webcam its only predicting as sad. what should i do?

What is the accuracy?

well explained but after training ,the model is not predicting well. only your model working fine. and brother stop copy and pasting the code. thanks

im not getting model.h5 file bro ... what to do please help bro

Handwritten signature and dates from image extraction can you do a video on this

Sir! for feature extraction which algorithm you have used?

THANK YOU!

Nice Explaination 👍

very good video thanks

model.h5 file is not appearing after running the code in kaggle. Give me solution for it bro.

Hi Akshit! Thank you for your helpful video. Can you explain more about the cascade classifier? if I want to detect other than face, can I use the same haarcascade xml file? thanks in advance.

I am getting error load_model is not defined in vscode please guide me

amazingggg

Do you have a 3D image??? End of Semester Project (3d-CNN) and thanks for the video

Can we use transfer learning on grayscale images?

Thanks for this video :)))

why is the frame rate so low?and how to improve the fps

Can you give some idea to web development using this ml algorithm (emotion detection)

How did you calculated that the accuracy is 72%?

How to run the code 😢help me pls. last step

Hey Akshit, I'm getting a Illegal instruction: 4 error. Do you know why? Thanks.

where to get the pretrained model from?

Bro module to import kr dete.. Kon sa module import kiy th.. Pip. Ke through

Also in your code I saw it took you a few seconds to calculate each epoch while my code is taking all the way to 12 minutes to calculate a single epoch? Could there be something wrong with my code?

How did you choose those layers? layer count, numbers of filters, etc (hyper parameters) please make a video!

Why there is no predicted value not labeled with predicted emotion?

kaggle part code possible to run in Google Colab?

bro i want to download the ipynb notebook that you posted in your github..

Helpful Video! why sometimes images are not display clearly?

Thanku..

how to choose the path in jupiter notebook.

bro how to change the size of openCV window?

Hi, thanks for your video. It's very helpful for my project. I have a question : I write the proportions of the emotions measured on the screen. When there is more than 1 person, the data is overwritten. How do I solve this?

Thanks for your clear and concise explanation. I'm stuck with a value error on MaxPooling 2D. can you post the exact code here?

i am not getting model.h5 file in output

do you have documentation for this sir?

If you can integrate a music player with this model then it will be a big help. Thank you

Currently, I am working on a pain detection project. This was very helpful, Thank you so much.

Are you post full clear code or not

How to build this with flutter

I have one dought instead of using our own weights can we use vgg16/19 to get better accuracy. In real world app will we use our own weights or predefined one?

which algorithm u have used ?

what is mode='max' doing

Thank you so much bro. U have explained very well. This is crystal clear for me. Actually i am beginner so i am going to try this for my mini project. Can u pls tell me that what are all packages i need to install for this . Plz tell me

Can u guide me to do this project

Can we put the program on heroku software by using python?! Will it be okay?

Hey! What is the version of tensorflow you are using, I'm getting errors.