The system uses different algorithms to provide face recognition, and face training (creation of invariant face data from multiple people’s images)
Face recognition solution. The main goal of the project is to create face-recognition system for the Retail industry. The system uses different algorithms to provide face recognition, and face training (creation of invariant face data from multiple people’s images)
Main features of the solution:
• Investigation and integration of auto-enrollment algorithm;
• Multiple image sources (including custom devices);
• high-performance image-processing routines (using IPP)
• intercommunication with other systems, e.g. video-tracking system;
PhotoButler is a US-based startup that sells an image recognition application that allows automatic creation of photo albums based on the content of users’ photos, using advanced machine learning techniques to achieve this.
Our team implemented Face Recognition module. We used Microsoft Face API as a recognition tool and developed a wrapper on top of it to integrate the module with the PhotoButler’s overall architecture.
We also performed analysis based on raw data obtained from users’ photos. The goal was to create albums of photos through building clusters of different occasions such as users’ birthdays, Christmas, Easter, Thanksgiving, Halloween, Hanukkah, etc.
Need a technological solution?