A Partially Binarized Hybrid Neural Network System for Low-Power and Resource Constrained Human Activity Recognition.
Antonio De VitaAlessandro RussoDanilo PauLuigi Di BenedettoAlfredo RubinoGian Domenico LicciardoPublished in: IEEE Trans. Circuits Syst. (2020)
Keyphrases
- resource constrained
- low power
- human activity recognition
- hybrid neural network
- activity recognition
- power consumption
- high speed
- low cost
- human activities
- resource constraints
- wireless sensor networks
- neural network
- security mechanisms
- sensor networks
- embedded systems
- rfid tags
- multipath
- logic circuits
- energy consumption
- low power consumption
- sensor nodes
- object oriented
- cmos technology