Neural sensor translation
Md Abu Sayeed Mondol, John A. Stankovic · 2019
Neural Machine Translation (NMT), a neural network based approach for language translation, has been shown to be more effective than traditional approaches like statistical machine translation. Inspired by NMT, we propose Neural Sensor Translation (NST), a process of translating data sequences from one or a set of sensors to another using neural networks. NST is a data-driven approach for creating virtual sensors that have useful applications in sensor networks and internet of things. We present a neural network based approach for sensor translation and demonstrate the potential of the approach using sensor data from a home.