CodeSquare

• 

  Home page where emotion is detected

• 

  Booklist for sad emotion

• 

  Booklist for happy emotion

Inspiration

The emotion-based song suggestion algorithm served as our source of inspiration. We believed that, similar to how an emotion-based song recommendation engine recommends songs, it could likewise suggest books that can inspire individuals.

What it does

The system will detect the emotion as user upload an image, and depending on that emotion, a list of books will be suggested.

How we built it

Python libraries such as tensorflow and keras are used in the backend , while html, CSS, and javascript are utilised in the front end. Finally, the flask framework was used to combine the backend and the front end.

Challenges we ran into

The biggest difficulty we encountered while working on this project was that although we hoped to identify 7 emotions, the accuracy was insufficient to provide an accurate result; as a result, we had to limit the emotions to only 2 emotions.

Accomplishments that we're proud of

We obtained an accuracy of about 80% after reducing the emotions to 2, which indicates that 8 out of every 10 users will receive a list of books that is accurate.

What we learnt

We learnt to train the model and how the emotion can be detected using the image as an input

What's next for CodeSquare

To develop a system that has an accuracy of about 90% for detecting seven emotions

Built With

• css3
• flask
• html5
• javascript
• python