A General Framework for Smoothing Arbitrary Signals in Computer Graphics and Biomedicine
Simone Cammarasana  1@  , Paolo Nicolardi  2  , Giuseppe Patanè  1  
1 : Istituto di Matematica Applicata e Tecnologie Informatiche
2 : Esaote S.p.A.

We propose a novel framework for the smoothing of arbitrary signals, which combines regularisation with learning-based models and is general with respect to the input signal, the noise type (e.g., speckle, Gaussian noise), the selected regulariser/denoising (e.g., SVD - Singular Values Decomposition, block matching), and the learning architecture (e.g., network's weights optimisation).

Personnes connectées : 3 Vie privée