From a fast-paced grasping football game of Bruins vs Trojans to that gold finish of Michael Phelphs, you pull out your phone to snap that victory moment only to find that you were a few seconds too late! Frustrated? Angry? Each of us has been in a similar position, but we decided to step forward and solve it once for all.
What it does
With Resurrect you can scroll back in time and re-capture those fragile moments. Just swipe horizontally to start a time travel, sourcing images from other users around you. Swipe vertically to see more images for a particular moment. There is a smart AI engine to provide the images that are relevant to the image just uploaded by you. Choose from the smart auto-generated captions and share away those pictures.
How we built it
Flask on Microsoft Azure as the backend. MongoDB for persistence storage. Azure App Service for Docker image Microsoft APIs 1) Computer Vision 2) Emotion 3) Face
Challenges we ran into
Getting Flask to run on Azure. Generating Thumbnails. Generating relevant quotes for a given image.
Accomplishments that we're proud of
Creating a smart engine to recommend relevant images to the user, thanks to Microsoft cognitive services!
What we learned
We learned the value of testing every component individually. This being a complex projects required all of us to work on multiple subparts, but it took us a plenty of time to debug when we integrated everything together, partly because the subparts were error-prone. We were delighted by the power of cloud computing, and the ability to harness intelligent APIs and get results within seconds. On technology side, we learned to use Microsoft Azure, and to integrate commonly used services like Flask with it.
What's next for Resurrect
Olympics 2024 360 degree view for Microsoft Holo Lens. Analytics for Event Organizers. Dynamic GeoFilters i.e. putting live scores of a match at the time a picture was taken.