Study of Noise Enhancing Sense Based on Psychophysical Method

Stochastic resonance phenomenon in a biological sensory system has been studied through signal detection theories and psychophysical experiments. In this paper, a sensory system is considered as a threshold detector including a receiver and a classifier. In comparison with conventional models regarding the receiver of sensory system as a linear or single non-linear system, a summing network is constructed by McCullochPitts (MP) neurons to simulate the receiver. The simulation results reveal that the relevant index of signal detectability exhibit the stochastic resonance characteristics. Theoretic results have been experimentally verified employed 2IFC (two interval two alternative forced choice) method. These works prove that stochastic resonance is not an epiphenomenon in sensory systems.