Partial Sum Minimization of Singular Values in RPCA for Low-Level Vision

===================================================================
                Graphic Group Seminar
===================================================================
Department of Computer Science and Engineering
Center of Visual Computing and Image Science
-------------------------------------------------------------------

Speaker:        Dr. Yu-Wing Tai
                Department of Electrical Engineering
                KAIST

Title:          "Partial Sum Minimization of Singular Values in RPCA
                 for Low-Level Vision"

Date:           Thursday, 21 November 2013

Time:           4:00pm - 5:00pm

Venue:          Room 1504 (near lifts 25/26), HKUST

Abstract:

Robust Principal Component Analysis (RPCA) via rank minimization is a
powerful tool for recovering underlying low-rank structure of clean data
corrupted with sparse noise/outliers. In many low-level vision problems,
not only it is known that the underlying structure of clean data is
low-rank, but the exact rank of clean data is also known. Yet, when
applying conventional rank minimization for those problems, the objective
function is formulated in a way that does not fully utilize a priori
target rank information about the problems. This observation motivates us
to investigate whether there is a better alternative solution when using
rank minimization.

In this work, instead of minimizing the nuclear norm, we propose to
minimize the partial sum of singular values. The proposed objective
function implicitly encourages the target rank constraint in rank
minimization. Our experimental analyses show that our approach performs
better than conventional rank minimization when the number of samples is
deficient, while the solutions obtained by the two approaches are almost
identical when the number of samples is more than sufficient. We apply our
approach to various low-level vision problems, e.g. high dynamic range
imaging, photometric stereo and image alignment, and show that our results
outperform those obtained by the conventional nuclear norm rank
minimization method. This work is published in the ICCV in this year.


****************
Biography:

Yu-Wing Tai is currently an assistant professor in the Department of
Electrical Engineering in KAIST. He received the BEng (first class honors)
and MS degrees in computer science from the Hong Kong University of
Science and Technology (HKUST) in 2003 and 2005 respectively, and the PhD
degree from the National University of Singapore (NUS) in 2009. He
received the Microsoft Research Asia Fellowship award in 2007, KAIST 40th
Anniversary Academic Award for Excellent Professor in 2011, and the Top 50
outstanding research projects supported by National Research Foundation
(NRF) in 2012 respectively. He regularly serves on the program committees
for the major Computer Vision conferences (ICCV, CVPR, and ECCV) and
reviewers for major journals including IEEE TPAMI, IJCV and IEEE TIP. His
research interests include computer vision and image processing.