The next generation of Internet of Things (IoT) is going to be dominated by Artificial Intelligence (AI) and Artificial Intelligence of Things (AIoT) is gaining popularity. The Artificial Intelligence and the deep learning models refined further can take the AIoT and the domains in which they are going to be used are going to see lot of advancements.
One of the simplest and widely used sensor and integrated with IoT is the motion detection sensor. Whether it be home, offices, motion sensor is widely used. What if we say, there is no need for motion sensor separately and based on the WiFi signal variations due to human movement, the motion can be sensed directly? Awesome, right? This is what researchers right now have introduced and this is called the multiple spectrogram fusion network (MSF-Net). MSF-Net is an advanced method for detecting the human movement inside the house. What happens in MSF-Net is that channel state information (CSI) of Wi-Fi is extracted to find the change of patterns and movement. The biggest challenge right now in motion sensing is movement of pets, other changes trigger false inputs causing unnecessary actions. With Wi-Fi signal variation based sensing, many problems might get resolved but requires a robust deep learning models to be developed.
]]>
0 Comments