scispace - formally typeset
J

Joe F. Chicharo

Researcher at University of Wollongong

Publications -  139
Citations -  2499

Joe F. Chicharo is an academic researcher from University of Wollongong. The author has contributed to research in topics: Adaptive filter & Blind signal separation. The author has an hindex of 24, co-authored 139 publications receiving 2382 citations. Previous affiliations of Joe F. Chicharo include Zhengzhou University & University of Oldenburg.

Papers
More filters
Proceedings ArticleDOI

De-cumulant based approaches for convolutive blind source separation

TL;DR: In this paper, the blind separation of signal sources (BSS) based on the approach of de-cumulant is studied and two algorithms are developed in the time-domain.

Special Session: Issues in Australian ICT Education

TL;DR: The Australian Software Engineering Conference is concerned with ICT education from the holistic perspective of preparation in high schools, the university experience, transition to the workforce, and the contribution by industry, government, and professional bodies as mentioned in this paper.
Proceedings ArticleDOI

Transition analysis for moderate feedback self-mixing interferometry

TL;DR: A theoretical analysis on the locations of transition points in moderate feedback self-mixing signal is presented in this article, where the locations for the start and end points for upward and downward switchings are calculated based on the Lang-Kobayashi model.
Book ChapterDOI

Joint diagonalization of power spectral density matrices for blind source separation of convolutive mixtures

TL;DR: A new approach for the blind separation of convolutive mixtures is proposed based on sources nonstationarity and the joint diagonalization of the output power spectral density matrices.
Journal Article

Joint diagonalization of power spectral density matrices for blind source separation of convolutive mixtures

TL;DR: In this paper, a blind separation of convolutive mixtures is proposed based on sources nonstationarity and the joint diagonalization of the output power spectral density matrices, which utilizes a time-domain separation network, but the coefficients are optimized based on a frequency domain objective function.