Poor Man's Rekognition

Rekognition is an Amazon service capable of identifying objects, text and activities, performing facial analysis and recognition, detecting the frequency of objects or an inappropriate scene, and much more using deep learning. This GSoC proposal aims to build a self-contained module to start off Poor Man’s Rekognition — an open-source version of the commercial service. This initial module, which will be scalable and robust, will have this as the main goal: “Given a set of properly tagged people (suppose for example, a number of celebrities), create an API that can be used to identify such people in other images.” The aim by the end of GSoC is to complete a pipeline that outputs a timeline of scenes in a video, and the different actors in each scene.