The idea is to build a mobile application in iOS and Android that lets you search for your look-a-like.
All you need to do is take a selfie and let the Facial Recognition algorithm do the work.
Initially, it would only be able to recognize and match those within the pre-model that have been indexed. If we utilize existing train models, it would provide a good start.
As of this writing, there is a pre-train model for matching celebrities. It would be easier to match a celebrity. But for a much larger pool, this would not be able to provide a good match until we have a larger registry.
If we add reinforcement learning to our Machine Learning algorithm, our models will get updated with new users and other sources such as Facebook, Instagram, etc.
Overtime our model will become more accurate with a larger pool to analyze against.
When new matches are found, matched users would be notified and provide them with details to connect if they choose to.
I am a Xamarin Developer and Founder of Xamo Toolkit. Xamo Toolkit is my way of helping everyone with a dream and basic app development know-how take their shot at the lucrative mobile app market.