Computer scientists at the University of Waterloo have created a smartphone app that guides users to capture the perfect selfie. Powered by a sophisticated algorithm, it provides real-time instructions on camera positioning for optimal results.
“Selfies have become a primary way for people to share their experiences, but quality varies widely,” explains the lead computer science professor. “Unlike post-processing apps, ours offers proactive guidance, helping users understand and improve their technique.”
To build the algorithm, researchers used 3D digital scans of average individuals. They simulated hundreds of virtual selfies by programming a virtual smartphone camera and computer-generated lighting, testing variables like light direction, face position, and size.
Through an online crowdsourcing platform, thousands voted on the best virtual selfies. Voting patterns were analyzed mathematically to refine the algorithm. In real-world tests, participants using the standard camera app versus the Waterloo app showed a 26% improvement in selfie quality, based on additional online ratings.
Watch a video here to see how this app works
