Tracking Of Human Activities

Abstract

Is it possible to know what is happening at a particular place precisely without seeing video clips where we are physically absent. Nowadays all of us being engaged with other works we don't have much time to spend on seeing the full-length video to know what is happening. But there is an alternative option for this, i.e in the form audio clip which is exactly like a person narrating the scene. The main advantage of this is we can simultaneously save time and multi-task i.e doing our work by listening to the audio clip that is generated by getting the up to date information and also if any person suddenly falls which may cause heavy injuries which can indirectly lead to a major medical issue for elderly people. So to prevent such emergencies, it will also provide an alarm system to detect human falls. This is possible by using current trending technologies like image processing and computer vision to capture the live moments, RNN with LSTMs to process and analyze the captured ones and by using natural language processing we can describe what is going on. Audio clips are generated by using Google Text to Speech API and they are sent to users.