Stochastic Systems Group
Home Research Group Members Programs  
Demos Calendar Publications Mission Statement Alumni

SSG Seminar Abstract


Stochastic Resonance in Signal and Image Processing

Pramod K. Varshney
Syracuse University


Stochastic resonance (SR) is a phenomenon in which the performance of some nonlinear systems can be enhanced by adding suitable noise under certain conditions. This counter-intuitive phenomenon has been observed in many fields such as Physics, Biology, and Neuroscience. The basic idea of performance enhancement by adding noise has been practiced in signal processing for some time, e.g., dithering in digital audio systems. We have recently formalized this idea and have developed the theory for SR in signal detection and estimation systems. We have obtained results on whether or not a system is improvable by SR and if yes, what the optimum noise probability density function is for the specific signal processing task. The results are quite surprising in that the form of the optimal noise is quite simple. This talk will introduce the phenomenon of SR, present our recent results and will conclude with application of SR to medical image processing.



Problems with this site should be emailed to jonesb@mit.edu