This is the preliminary stage of a machine speech+ vision system. Jarvis (yes I’m an ironman fan) is the AI speech bot which can understand commands (and continuous speech as well) and react to it with its abilities. The vision system can track any known or unknown objects and learn about them and keeps a record of the learnt objects. A sample is shown in this video. It can track and learn multiple objects simultaneously ( see my other video here :http://www.youtube.com/watch?v=K59YRUS7v3A) . The system learns and tracks online (no offline descriptors) and can remember the objects for future tracking.