Inspiration
Sentiments are Something that Every Human Being Has and Should be Valued but We Live in a world where Languages Spoken are in Plenty and it is Difficult to Know all the Languages . Though there are many so many translators , we cannot use it very often to Understand and translate everything . So there is Something Needed that Would Make people from a different Language Base Understand and Value the Sentiments of a Person from a totally different Language Clan . There are nearly 6500 Languages around the World Especially in a Country Like INDIA where over 22 languages are spoken officially with so much diversity , Everyone cannot understand what you are trying to convey over mobile phone or it is very difficult to understand the Sentiments and Moods of Different people with different Languages and that is where Our Project Comes into play by showing the accurate Sentiment of people through their speech or even texts in any Languages .When a Phone Call is Run or an Audio is Played . Our Model will Display the Respective Sentiment of the Speaker so that the Receiver can Understand Easily .
What it does
The Speaker can speak any language and the receiver ( who does or does not know the language of the speaker ) will be able to understand the tone , sentiments of the speaker through our model .
Our Model , Will get the Sentiment of a Person through his Voice or Speech or Any Audio File and of Any Language . It takes the Audio File as Input from the User and Converts Into text and then and it goes through the translation model . After It has been Converted to English . We Run our Sentiment Analyzing Model and Display the Sentiment of the Speaker through an Emoji . An Emoji that depicts the Respective Sentiment is Shown as Output to the User.
How I built it
For the User-Interface that is the Front-end we have used Html , CSS and Javascript for the Webpage . The Entire Model was developed Using Python . For Capturing the Speech we Used PyAudio Module and Speech Recognition Model . The Speech was Converted to text, after that We Used the Python Module to translate the given language to English using Natural Language Processing . Then the text was sent to the Sentiment Classifier Model and the respective Sentiment was found according to the speech of the speaker . After the Sentiment was Classified . The Output was displayed through Emojis for respective Sentiment using python's emoji module. The code was then deployed Using Flask . The Website was created Using Flask. Flask is a lightweight Python web framework that provides useful tools and features that make creating web applications .
What I learned
Learnt to Make Sentiment Classifier Model , Speech Recognition Model . Learnt to Make Web applications Using Flask .
What's next for Multilingual Sentiment Analyser
Will be Build as a Web App by Us and Can be Used parallelly when a Phone call is Run or When a Audio is Played. So that Receiver can get to know the Sentiment of the Speaker in Emoji Form .
Can be Used in Customer Feedback Centre's , Where Call Centre Employees get Easily understand the sentiments of the Customer Via Phone Call .
Now , any Person in the World can easily understand the Mood or the Sentiment of Any other Person with any other Language through his Speech or Voice .
Log in or sign up for Devpost to join the conversation.