An incremental principal component pursuit algorithm via projections onto the ℓ1 ball

Video background modeling, used to detect moving objects in digital videos, is a ubiquitous pre-processing step in computer vision applications. Principal Component Pursuit (PCP) PCP is among the leading methods for this problem. In this paper we proposed a new convex formulation for PCP, substituting the standard ℓ1 regularization with a projection onto the ℓ1-ball. This formulation offers an advantage over the known incremental PCP methods in practical parameter selection and ghosting suppression, while retaining the ability to be implemented in a fully incremental fashion, keeping all the desired properties related to such PCP methods (low memory footprint, adaptation to changes in the background, computational complexity that allows online processing).

[1]  Brendt Wohlberg,et al.  A Matlab implementation of a fast incremental principal component pursuit algorithm for Video Background Modeling , 2014, 2014 IEEE International Conference on Image Processing (ICIP).

[2]  Brendt Wohlberg,et al.  Incremental Principal Component Pursuit for Video Background Modeling , 2015, Journal of Mathematical Imaging and Vision.

[3]  M. Brand,et al.  Fast low-rank modifications of the thin singular value decomposition , 2006 .

[4]  Soon Ki Jung,et al.  Decomposition into Low-rank plus Additive Matrices for Background/Foreground Separation: A Review for a Comparative Evaluation with a Large-Scale Dataset , 2015, Comput. Sci. Rev..

[5]  Namrata Vaswani,et al.  Support-Predicted Modified-CS for recursive robust principal components' Pursuit , 2011, 2011 IEEE International Symposium on Information Theory Proceedings.

[6]  Martin Kleinsteuber,et al.  pROST : A Smoothed Lp-norm Robust Online Subspace Tracking Method for Realtime Background Subtraction in Video , 2013, ArXiv.

[7]  Yi Ma,et al.  The Augmented Lagrange Multiplier Method for Exact Recovery of Corrupted Low-Rank Matrices , 2010, Journal of structural biology.

[8]  Namrata Vaswani,et al.  An Online Algorithm for Separating Sparse and Low-Dimensional Signal Sequences From Their Sum , 2013, IEEE Transactions on Signal Processing.

[9]  James C. Bezdek,et al.  Some Notes on Alternating Optimization , 2002, AFSS.

[10]  Thierry Bouwmans,et al.  Robust PCA via Principal Component Pursuit: A review for a comparative evaluation in video surveillance , 2014, Comput. Vis. Image Underst..

[11]  John Wright,et al.  Robust Principal Component Analysis: Exact Recovery of Corrupted Low-Rank Matrices via Convex Optimization , 2009, NIPS.

[12]  Brendt Wohlberg,et al.  Incremental Principal Component Pursuit for Video Background Modeling , 2015, Journal of Mathematical Imaging and Vision.

[13]  Laura Balzano,et al.  Incremental gradient on the Grassmannian for online foreground and background separation in subsampled video , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[14]  Paul Rodríguez,et al.  An accelerated newton's method for projections onto the ℓ1-ball , 2017, 2017 IEEE 27th International Workshop on Machine Learning for Signal Processing (MLSP).

[15]  Yoram Singer,et al.  Efficient projections onto the l1-ball for learning in high dimensions , 2008, ICML '08.

[16]  G. Sapiro,et al.  A collaborative framework for 3D alignment and classification of heterogeneous subvolumes in cryo-electron tomography. , 2013, Journal of structural biology.

[17]  Brendt Wohlberg,et al.  Fast principal component pursuit via alternating minimization , 2013, 2013 IEEE International Conference on Image Processing.

[18]  Laurent Condat,et al.  Fast projection onto the simplex and the $$\pmb {l}_\mathbf {1}$$l1 ball , 2016, Math. Program..

[19]  Fatih Murat Porikli,et al.  CDnet 2014: An Expanded Change Detection Benchmark Dataset , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops.

[20]  Brendt Wohlberg,et al.  Ghosting suppression for incremental principal component pursuit algorithms , 2016, 2016 IEEE Global Conference on Signal and Information Processing (GlobalSIP).

[21]  CondatLaurent Fast projection onto the simplex and the $$\pmb {l}_\mathbf {1}$$l1 ball , 2016 .

[22]  Yi Ma,et al.  Robust principal component analysis? , 2009, JACM.

[23]  James M. Rehg,et al.  GOSUS: Grassmannian Online Subspace Updates with Structured-Sparsity , 2013, 2013 IEEE International Conference on Computer Vision.

[24]  Yong Yu,et al.  Robust Subspace Segmentation by Low-Rank Representation , 2010, ICML.

[25]  Paul L. Rosin Unimodal thresholding , 2001, Pattern Recognit..

[26]  Brendt Wohlberg,et al.  Video background modeling under impulse noise , 2014, 2014 IEEE International Conference on Image Processing (ICIP).