Open rashmisgh opened 6 years ago
@rashmisgh you can calculate embedding of each face and append to a list of dict then calculate embedding of a face from webcam then calculate distance of this from each embedding from the list. It should work for thousands of persons within in 1 second but is not scaleable. for more persons you can use classifier. You can pickle list to file for later use.
Hey can you implement this recognition through live camera? The user input will be a image, and the feature is stored in a vector and will running the live camera the same face get recognised.