Inspiration

We were originally considering creating an application that would take a large amount of text and summarize it using natural language processing. As well, Shirley Wang feels an awkward obligation to incorporate IBM's Watson into the project. As a result, we came up with the concept of putting in an image and getting a summary from the Wikipedia article.

What it does

You can input the url of a picture into the webapp, and it will return a brief summary in bullet point form of the Wikipedia article on the object identified within the picture.

How we built it

We considered originally using Android Studio, but then a lot of problems occurred trying to make the software work with it, so we switched over to Google App Engine. Then we used Python to build the underlaying logic, along with using IBM's Watson to identify and classify photos, the Wikipedia API to get information from Wikipedia articles, and Google's Natural Language API to get only the key sentences and shorten them down to bullet point form while maintaining its original meaning.

Challenges we ran into

We spent over 2 hours trying to fix a Google Account Authentication problem, which occurred because we didn't know how to properly write the path to a file, and Pycharm running apps is different from Pycharm running apps in its own terminal. We also spent another 2 hours trying to deploy the app, because Pycharm had a screwed up import statement and requirements file that messed up a lot of it.

Accomplishments that we're proud of

This is our first hackathon and our first time creating a web app, and we're really happy that we managed to actually successfully create something that works.

What I learned

Sometimes reading the API carefully will save you over half of your debugging time in the long run.

What's next for Image Summarizer

Maybe we'll be able to make a way for the users to input a photo directly from their camera or their computer saved photos.

Built With

Share this project:

Updates