eprintid: 3703
rev_number: 7
eprint_status: archive
userid: 17
dir: disk0/00/00/37/03
datestamp: 2019-12-04 07:33:25
lastmod: 2019-12-04 07:33:25
status_changed: 2019-12-04 07:33:25
type: conference_item
metadata_visibility: show
creators_name: Le, Trung Thanh
creators_name: Nguyen, Viet Dung
creators_name: Nguyen, Linh Trung
creators_name: Karim, Abed Meraim
creators_id: letrungthanhtbt@gmail.com
creators_id: nvdung@vnu.edu.vn
creators_id: linhtrung@vnu.edu.vn
creators_id: karim.abed-meraim@univ-orleans.fr
title: Robust subspace tracking with missing data and outliers via ADMM
ispublished: pub
subjects: ECE
subjects: IT
divisions: avitech
divisions: fac_fet
abstract: Robust subspace tracking is crucial when dealing with data in the presence of both outliers and missing observations. In this paper, we propose a new algorithm, namely PETRELS-ADMM, to improve performance of subspace tracking in such scenarios. Outliers residing in the observed data are first detected in an efficient way and removed by the alternating direction method of multipliers (ADMM) solver. The underlying subspace is then updated by the algorithm of parallel estimation and tracking by recursive least squares (PETRELS) in which each row of the subspace matrix was estimated in parallel. Based on PETRELS-ADMM, we also derive an efficient way for robust matrix completion. Performance studies show the superiority of PETRELS-ADMM as compared to the state-ofthe-art algorithms. We also illustrate its effectiveness for the application of background-foreground separation.
date: 2019-09
date_type: published
contact_email: linhtrung@vnu.edu.vn
full_text_status: none
pres_type: paper
event_title: 27th European Signal Processing Conference (EUSIPCO)
event_location: Coruna, Spain
event_dates: September, 2019
event_type: conference
refereed: TRUE
citation:   Le, Trung Thanh and Nguyen, Viet Dung and Nguyen, Linh Trung and Karim, Abed Meraim  (2019) Robust subspace tracking with missing data and outliers via ADMM.  In: 27th European Signal Processing Conference (EUSIPCO), September, 2019, Coruna, Spain.