Floo

• 

• 

  Pop up with flight information

Inspiration

When thinking of places to travel to, we wanted to include places we've seen in movies or content we've come across while surfing the internet. We wanted to create a project that could help people discover locations and book flights on JetBlue, and participate in experiential travel based on the locations featured in their favorite media and entertainment. Just like the Floo network of Harry Potter, users could now surf the world while surfing the web.

What it does

Floo is a Google Chrome Extension that enables users to take a screenshot of their browser content, whether it be a video they're streaming or image on a webpage. Floo then recognizes what location in the world the image taken, and searches for JetBlue flights to that destination. Our extension displays alerts to the user that show the cheapest flights and hottest deals for the location captured in the image, and invites them to make a booking.

How we built it

Throughout our frontend and backend, we utilized Google technology, and loved its usage in our unique task of location detection. We used HTML, Javascript, and the chrome.extension API to build the extension. For our backend, we used the Google Cloud Vision API's Landmark Detection and Web Detection models, which determined the place depicted in our image by searching the web for similar images and identifying landmarks. We then used Google Maps API to convert geographic coordinates to addresses, from which we found the closest airport to our location serviced by JetBlue. Lastly, we parsed the JetBlue low fares and deals CSV data, and returned the cheapest flights from our origin to the destination. We used Flask as the server to interface between our frontend and backend.

Challenges we ran into

There were still limitations as to the types of places computer vision models performing location detection could accurately recognize. In addition, we ran into trouble sending POST requests in Javascript, and combining our front end with our back end.

Accomplishments that we're proud of

Creating a project that could naturally fit in with the user's browsing experience, and finding accurate flight results that could a.

What we learned

Google Cloud Vision API, Maps API, Chrome Extension, Flask, Javascript, HTML, Python

What's next for Floo

In the future we hope for Floo to be able to generate suggested locations even from generic locations. It could achieve this by analyzing key features from the image (e.g. fall foliage, mountains, beachside), and searching for JetBlue destinations that matches these descriptions.