In collaboration with Nodeflux, we are working to develop an automatic system for detecting room activity. This system utilizes video captured by CCTV inside a room. There are various benefits that can be obtained from this system. For instance, it can be used to optimize room scheduling in real time. It also can be employed to adjust any peripheral in the room to save power.
This research is also contributes to the body of knowledge by proposing novel deep learning algorithm for object counting. The proposed algorithm combines the concept of attention mechanism and neural arithmetic logic units in a creative way. This combination is aimed to enable the developed system to detect room activity level.
Published at : Updated