Lazy Captions

Inspiration

Just having a great picture to upload on facebook or Instagram is not sufficient. It has to be uploaded with a great caption too. I have seen people searching all over the internet for that one perfect caption. We decided to solve this problem.

What it does

Lazy caption takes a picture from a user.......abbbrraacaadabraaa and you get the perfect caption.

How we built it

Lazy caption works on object detection. We detect objects in the picture, process their probabilities and get search for the caption in our database.

Challenges we ran into

Which objects to select, which quotes to put in the database. Pictures having one or more objects.

What we learned

Getting our hands dirty with IOS and Android permissions. Setting up a fancy search algorithm in AWS CloudSearch.

What's next for Lazy Captions

To make caption suggestions more and more relevant.

Built With

Submitted to

HackBU

Created by

I worked on the Backend.
Wrote the server in Python Flask and deployed on Heroku.

Akshay Borse
Established AWS cloudsearch and wrote a python script to create file that needs to be uploaded for database that holds captions. Did web scraping for the quotes.

Nehal Pawar
Worked on design, development and integration of react-native app.

Shashank Kaldate
Priyanka Hiremath

Updates

Akshay Borse started this project — Feb 10, 2019 12:28 PM EST

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.