Inspiration
Sojourn started with an amazing experience with Pokemon Go. After endless nights of late night exploring and wasting all of our pokeballs, we wondered how we could make people explore their surroundings through the framework of a smartphone application, and also track their calories while doing so!
What it does
Sojourn randomly selects nearby locations and pulls an image on the ground through the use of Google Streetview. Then you are invited to participate in a "quest" to travel to that location, timing you and recording fitness data. Once you reach that place, you must take a picture in the same perspective as the Google Streetview picture. Using Machine Learning, it compares the two images and verifies that you have completed the quest!
How we built it
We used a RESTful design to implement a back-end Node.js server and a front-end IOS application, as well as using the Google Streetview API to dynamically pull nearby streetviews. We were able to include compass integration with the map and autorotation with phone orientation, and use ML libraries.
Challenges I ran into
We were struggling to find ways to import Google Streetview images into the Node.js server, as well as implementing machine learning algorithms in comparing the images.
Accomplishments that we are proud of
We got it to work!
What we learned
Node.js, implementing Streetview, and various quirks of Swift libraries.
What's next for Sojourn
We are trying to implement Google Fit's API so that users can both import and export fitness profiles between fitness application. In addition, we plan to implement cloud scalability to offload the computational load from the smartphone onto services such as AWS.
Log in or sign up for Devpost to join the conversation.