A stochastic diffusion approach to signal denoising

We present a stochastic formulation of a linear diffusion equation (or heat equation), and in light of the potential applications ranging from signal denoising to image enhancement/segmentation of its nonlinear extensions, we propose a more general nonlinear stochastic diffusion. The constructed stochastic framework, in contrast to traditional deterministic approaches, unveils the sources of of existing limitations and allows us to further significantly improve the performance by addressing the key problem. Substantiating examples are provided