-

 

Face Emotion Recognition

Introduction :-    

While digital platforms have limitations in terms of physical surveillance , it comes with the power of data and machines , which can work for you .

It provides data in form of images , video , audio and texts , which can be analyzed using deep learning algorithms . A deep learning -backed system not only solves the surveillance issue , but also removes the human bias from the system , and all information is no longer in the teacher's brain but rather translated into numbers that can be analyzed and tracked . 


Problem Statements :-

We aim to solve one of the challenges faced by digital platforms by applying deep learning algorithms to live Images Data .

We do this by recognizing the facial emotion of the participants using the CNN model we create will categorize the observed emotion accordingly.


DATA Set Used:-


We have utilized the FER 2013 data sets provided on Kaggle.

The data consists of 48*48 pixels grayscale images of Faces .


Data Pre-Processing:-


 What Happens inside an CNN ?




Uing ResNet-50 Architecture?

ResNet-50 is a Convolutional neural network that is 50 layers deep . ResNet  , short for Residual Networks is a classic neural network used as a backbone for many computer vision tasks .

The fundamental break through with ResNet was it allowed us to train extremely deep neural network with 150+ layers . It is an innovative neural networks that was first introduced by Kaiming He , Xiangyu Zhang , Shaoqing Ren , and Jian Sun in their 2015 Computer vision research paper titled Deep Residual learning for Images Recognition .

Convolutional Neural Networks have a major disadvantages - 'Vanishings Gradients Problems' . During backpropagation the values of gradients decreases significantly , thus hardly any changes come to weights . To overcomes this , ResNet is used  it make use of  'SKIP CONNECTION'

Plots for Accuracy And Loss :

The created model was run for 30 epochs in order to get an accuracy of 69.24% on training set and 62.50% on validation set.


Real Time Facial Emotion Detection :

It was also capable of detecting multiple faces and their respective emotions successfully . The face detection was done using OpenCV , and the output was displayed on the webpage using Streamlit packages.


Summary :-

We Successfully achieved our objective and created a model capable of recognizing seven different classes of emotion mentioned in the dataset. For it ti be accessible by others , a web app was created using OpenCV and Streamlit Library.

Both locally deployed app and the webapp can recognize the facial emotions of a user using the webcam    of his/her laptop or computer.


Requirement:- Pycharm - Python

                        Local Deploy - Streamlit 

                           Cloud Deploy - AWS

Files :-https://drive.google.com/file/d/1XLKWwPWJQcnebEWQL3waTkQiHiPf2wwc/view?usp=sharing

Comments

Post a Comment

Popular posts from this blog

-
Image
Prediction using Supervised ML First Question comes to mind What is the Supervised Machine Learning (ML) ? Supervised learning, also known as supervised machine learning, is a subcategory of  machine learning  and  artificial intelligence .  It is defined by its use of labeled datasets to train algorithms that to classify data or predict outcomes accurately. As input data is fed into the model, it adjusts its weights until the model has been fitted appropriately, which occurs as part of the cross validation process. Supervised learning helps organizations solve for a variety of real-world problems at scale, such as classifying spam in a separate folder from your inbox. Example of Supervised Learning Algorithms:    Linear Regression Logistic Regression Nearest Neighbor Gaussian Naive Bayes Decision Trees Support Vector Machine (SVM) Random Forest So We are using the Linear Regression for the Assignment . So What is Linear Regression ? Linear regression is u...
Read more