Emotion Detection from Text and Media

When social media was born, people started using it. their communication in face to face became less, more lesser then before the pendipic came. people become less emotional. because they don?t know the emotion behind the text messages. So, we decided to make a web-based chatting applic

Project Title

Emotion Detection from Text and Media

Project Area of Specialization

Artificial Intelligence

Project Summary

When social media was born, people started using it. their communication in face to face became less, more lesser then before the pendipic came. people become less emotional.
because they don’t know the emotion behind the text messages. So, we decided to make a
web-based chatting application that helps people to detect their emotions and the others
while they communicate with each other to make the emotional gap more less.

Emotion detection from media and text is web-based chatting application that the users
should first login if they are not already logged in. The users must log in with their Google
account. After successful login the users will be directed to the main page of the web
application.And can logout by clicking to his/her profile picture which is on the most top
left side of the application and directed to the login page.
If the user is logged in the first time, there will be no list of chats then s/he can create a new
chat by clicking either to the button under the search bar or the “chat” icon above it. then
the user can enter the email of any user he wants to chat with. The email address will
appear, profile picture and last message if exit on the sidebar of the application. When the
user clicks on the chat s/he created the chat screen will appear on the right side of the
application which takes 60% of the width of the screen and another 40% for the sidebar in
the left side of the application. Then s/he can start chatting with other users by sending the
text, voice recording, and image. Also can make audio calls and video chat.
We are using Google Firebase website; which handles the authentication and storing the
database into it as the backend of our application, and as the frontend we are deploying
automatically into the Vercel website whenever we push our code into the main branch of
our repository in Github.
Whenever the users text, voice, and video chat with each other the data being stored into
the database will be analyzed using machine learning and deep learning to detect the
emotion of the users and tell to the person they are chatting with. The text chat will analyze
and display the output of the emotion on the chat screen next to the where the text chat was
displayed, the voice recorded will be converted to text and then analyzed same like the text
chat. Also, the video chat will analyze using deepface which detects the face and tells the
emotion of the person while the video chat is going on.

Project Objectives

A web-based chatting application that helps people to detect their emotions and the others
While they communicate with each other to make the emotional gap less. With the help of technology, we are going to help the person suffering from psychological diseases like depression and anxiety.

These are our main objectives of this project.
1) To create a web application to allow the user to sign in with their google account.
2) To create a web application to allow the user to sign out from the application.
3) To create a web application to communicate in one to one.
4) To create a web application to allow users to create a new chat with anyone
through their email address.
5) To create a web application to allow users to send text to each other and to analyse
the message and to tell what is his/her emotion when they send it.
6) To create a web application to allow users to send recorded voices to each other
and to analyze the voice which will be converted to text first and to detect the
emotion of the person.
7) To create a web application to allow users to chat with voice call, and video chat.
8) To create a web application to video chat with users to see the live emotion
detection of the person’s face they are chatting with.

Project Implementation Method

We are using Google Firebase website; which handles the authentication and storing the database into it as the backend of our application, and as the frontend we are deploying automatically into the Vercel website whenever we push our code into the main branch of our repository in Github.
Whenever the users text, voice, and video chat with each other, the data being stored into the database will be analyzed using machine learning and deep learning to detect the emotion of the users and tell it to the person they are chatting with. The text chat will analyze and display the output of the emotion on the chat screen next to the where the text chat was displayed; the voice recorded will be converted to text and then analyzed same as the text chat. Also, the video chat will analyze using deepface which detects the face and tells the emotion of the person while the video chat is going on.

Tools:
? VSCode
? Firebase
? Github
? Vercel
Technologies
? Languages
? HTML/CSS
? JavaScript
? Python
? Frameworks

? Flask
? ReactJs
? NextJs
? PeerJs

Text Emoticons Detections:
In text, emotions are detected by the process of NLP and we use a pre-build library which
is known as text emotion.In Natural language processing we have multiple techniques to
clean data by useless word like”prepositions, articles, punctuations,spaces and so on”
that's why we normalize ,and tokenization.After cleaning we have step to use to train
predictive model like ”SVM,KNN, Naive Byse,and so on”.then we test the models by
comparing technique.in final state we use text2emotion library to detect emotion and
dominant the one emotion out of 5.which help to detect effective and efficient emotion
with accuracy above the 90%.

voice or speech recognition:

there is simple conversion of audio to text and the text to emotion with our Train model.

Benefits of the Project

Motive: Negative emotions into positive emotions

Emotions play a very important role in our mental health, we can see the pain, happyness, joy,
and fear through gestures or moods like smiling with happiness and a sad face that shows the sadness or pain. It’s hard to express your own gestures in text, even phone calls but in live video
call it's possible. That’s why we introduce a new technology which shows the right and correct
emotion of a person from text, voice, and video chat.
Emotions are likely for a minute but mood is for hours or a day, it depends on the feeling you feel
and the situation. Moods are not unique and nonverbal expressions but Emotions are universal
facial expressions.
Its human biological nature. If someone is in distress, we feel and understand the problem so we
try to resolve it and feel his comfort. Basically distress is a negative mood of emotion and we try
to change it in a comfort mood of emotion which is positive. On the other hand what's you feel
during helping and before it? you feel empathy “an affective response in which a person
understands and feels the distress of others” empathy creates other emotions like sympathy,
compassion, and tenderness.Psychologists and psychiatrists can easily find out the emotions and mood of other people but it's
hard to find for a normal person. It's not a problem to detect the mood but it's a problem to change
the negative mood to positive. Everyone faces the disorder on a daily basis but we cope up
because it's not our focus. A disorder deliberately focuses then it may cause of problem.like
depression, bipolar disorder, schizophrenia, anxiety, and eating disorder.
Through chat, our application can easily tell our emotions and emotions of people we are
communicating with. So, the self awareness a person can manage himself or another person to
help him to change into positive emotions.

Technical Details of Final Deliverable

The user end information and Guidline.

All the features of this project are:
? Login
Every user must login to use the application. When the user visits our website first will be
directed to the login page, if the user is not logged in already. In the login page there will
be a [SIGN IN WITH GOOGLE] button. When the user clicks a pop window will appear
and the user must enter his/her email and password of his Google account. Then the
information like email address and profile picture will be stored in the database. After
successful login automatically the user will be directed to the main page.
? Logout
If the user wants to logout from the web s/he can click on their profile picture and the
web will be directed to the login page.
? Create new chat
The user can create a new chat by clicking the [START A NEW CHAT] button or
“Message” icon. Then a prompt alert will popup and must enter the email address of the
person s/he wants to chat with. And then the email of the receipt and the user’s email will
be stored into the database.
? Search a chat
The user can search any chat from the chat list s/he has been created using the search bar.
when the user types the email address it will display that chat only where chats list ware
appeared.
? Send text
The users can send a text message through the receipt email, and the message will be sent
only to that email address and both the sender and the recipient can only read the
messages. The messages will be stored in the database and will be fetched to display on
the chat screen.
? Send recorded voice
A recorded voice message can be sent by clicking the “mic” icon either directly recorded
or inserted from the local device.
? Send image
The user can send any image that s/he inserted from the local device and stored to the
database and displayed on the chat screen.
? Send emoji
An emojis can be sent by clicking “WIN + dot” and sent as a text message.
? Voice call
A voice call can be made by any user to anyone in his/her chat list. We are using PeerJS
to implement this functionality.
? Video call
The users will be able to make a video chat to anyone they are chatting with. The video
will be analysed while the video chat is going on and the face will be detected and and analysed the emotion of the users and the dominant emotion will be displayed instantly as
detected the emotion change and each user will see the emotion of the other one.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Health

Other Industries

Media , Others , Telecommunication

Core Technology

Artificial Intelligence(AI)

Other Technologies

Others

Sustainable Development Goals

Good Health and Well-Being for People

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com