EmoDiary

• 

• 

  baisc text

• 

  color change according to emotion

• 

  after download

Inspiration

I always liked having a record of my life, So reading books, journaling have been an integral part of my life. What if there was an interesting way to entry diary or maintain a journal. Most of us have stopped decorating journals but I tried a different approach so one can play and learn.

What it does

there are as far two unique features. It is an easy navigating platform. For every 25 seconds in the dairy entry, the transformer model verifies the text and determine the sentiment. For every emotion recognised, the diary will change its color. I have used sample entries to test this. It also will have an unique way of adding stickers which is through ASL.

How we built it

we built the front end through HTML, CSS and Javascript. To work on the camera, sign language. and sentiment recognition, we used NLP and python. Also libraries like flask, transformers were used. Pre-trained model bhadresh-savani/bert-base-uncased-emotion was also used.

Challenges we ran into

The model prediction was not correct. Despite using a pretrained model. The accuracy was not correct. Also the laptop was quite burdened with the working of code. Also attempt was done but training an ASL model was not quite successful.

Accomplishments that we're proud of

using sentiment analysis to change the color of the page as per the emotion .

What we learned

Integrating languages to built multiple features. Color theory was being focused on. I learned the importance of understanding sentiment through text. our emotion is what makes us human.

What's next for EmoDiary

For EmoDiary next plan is to use an ASL model to integrate customized stickers. more features of decorations are to be added to make the diary entry more fun.

Built With

• css
• flask
• html
• javascript
• natural-language-processing
• numpy
• python
• transformers