Fundamental component enhancement via adaptive nonlinear activation functions

DOI10.1016/J.ACHA.2022.11.007MaRDI QIDQ2677841zbMATH OpenOpenAlexFDO

Authors Stefan Steinerberger, H.-T. Wu

Publication date 6 January 2023

Published in Applied and Computational Harmonic Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2112.01668

zbMATH Keywords

activation functionfundamental componentrectification

Mathematics Subject Classification ID

Signal theory (characterization, reconstruction, filtering, etc.) (94A12)

Abstract: In many real world oscillatory signals, the fundamental component of a signal $f(t)$ might be weak or does not exist. This makes it difficult to estimate the instantaneous frequency of the signal. Traditionally, researchers apply the rectification trick, working with $|f(t)|$ or $mboxReLu(f(t))$ instead, to enhance the fundamental component. This raises an interesting question: what type of nonlinear function $g:mathbbRightarrowmathbbR$ has the property that $g(f(t))$ has a more pronounced fundamental frequency? $g(t)=|t|$ and $g(t)=mboxReLu(t)$ seem to work well in practice; we propose a variant of $g(t)=1/(1-|t|)$ and provide a theoretical guarantee. Several simulated signals and real signals are analyzed to demonstrate the performance of the proposed solution.

Recommendations

• Time-frequency representation of nonstationary signals: the IMFogram
• Numerical instability of deconvolution operation via block pulse functions
• The role of functional data analysis for instantaneous frequency estimation
• scientific article; zbMATH DE number 1217250
• Weak convergence and boundedness properties of measures in frequency analysis

Cites work

• Decomposing Non-Stationary Signals With Time-Varying Wave-Shape Functions
• Extracting a shape function for a signal with intra-wave frequency modulation
• scientific article; zbMATH DE number 3148545 (Why is no real title available?)
• Instantaneous frequency and wave shape functions. I
• Periodicity transforms.
• Recursive diffeomorphism-based regression for shape functions
• Wave-shape function analysis. When cepstrum meets time-frequency analysis

Describes a project that uses

Uses Software

• PhysioToolkit

This page was built for publication: Fundamental component enhancement via adaptive nonlinear activation functions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2677841)