Self-organised attendees in festivals and events who don’t want to be photographed, opt out by putting a red dot on their forehead link , link. Also parents who upload pictures of their child in social media often blur the face or replace it with a playful emoji link.
What it does
Take a picture or load it from the gallery and blur underage faces. To convey the affect state of the kids in the photo, the face can be replaced with an emoji which corresponds to their current emotion.
How we built it
Patronus is an Android mobile application. Age and emotion are estimated through the Microsoft’s Project Oxford Face API, based on recent deep neural network research advances link, while blurring and emoji-based processing were handled with Java.
Challenges we ran into
We had to develop a fully functional MVP mobile app in a couple of hours, supporting the capability of remote API endpoint calls and local image processing tasks. The allocation of tasks and collaboration on shared code was challenging since we had to test the android application on many different phones. Framing the problem as a possible use case for parents and privacy-aware people in general led to interesting user experience choices such as putting substance over form (supporting two main use modes).
Accomplishments that we're proud of
In a matter of a few hours, we managed to come up with a socially relevant topic and to deliver a working prototype that achieves high levels of performance on the tasks of age estimation and emotion recognition. The excellent collaboration and understanding among the members of the team was crucial to that end.
What we learned
Facing the limited time of a 40-hour hackathon and the pressure to manage and create something that will be functional and also useful to other people really brings out the best in us. We like to call it creative anxiety.
What's next for Patronus
Refine the project into a product and approach the market with a suitable business plan.