Proximity Capacitive Gesture Recognition for Recursive Neighbor Memory Neural Network

This manuscript presents a proximity capacitive sensor that utilizes the Recursive Neighbor Memory Neural Network (RNMNN) for the detection of user gestures. The examination of human interactive gesture signals represents a significant research domain within the realm of smart home technology. In this context, algorithms are deployed on local devices to facilitate real-time recognition. Neural networks have found widespread application in various domains, encompassing tasks such as identification, control, and classification. The distinctive feature of RNMNN lies in its multi-recursive weights that facilitate inter-neuron communication, enabling the recording of neighboring neuron states to enhance global training performance. Moreover, the adoption of RNMNN-based proximity capacitive sensors offers several advantages, including heightened accuracy and robustness in discerning user gestures. The sensor's real-time detection capabilities, coupled with its capacity to adapt and learn from previous signals, position it as a promising solution for augmenting human-computer interaction and user experience within smart home environments. The integration of this approach into diverse smart home devices holds significant potential for propelling the field of human-computer interaction and elevating the overall functionality of smart homes.