Resonance-based Signal Analysis
Numerous signals arising from physiological and physical processes are not only non-stationary but also posses a mixture of sustained oscillations and non-oscillatory transients that are difficult to disentangle by linear methods. Examples of such signals include speech, biomedical and geophysical signals. This talk describes the decomposition of such signals into ‘resonance’ components: A high-resonance signal is one in which oscillations are sustained; while a low-resonance signal is one comprised mostly of non-oscillatory transients of unspecified shape and duration. While frequency components are straightforwardly defined and can be obtained by linear filtering, resonance components are more difficult to define and procedures to obtain resonance components are necessarily nonlinear.
The decomposition algorithm presented in this talk utilizes recent developments in signal processing, including sparse signal representations using SALSA and constant-Q (wavelet) transforms with tunable Q-factors.