RASL: Robust Alignment by Sparse and Low-Rank Decomposition for Linearly Correlated Images

Authors:
Yigang Peng;Arvind Ganesh;John Wright;Wenli Xu;Yi Ma
Affiliations:
Tsinghua University, Beijing;University of Illinois at Urbana-Champaign, Urbana;Columbia University;Tsinghua University, Beijing;University of Illinois at Urbana-Champaign, Urbana and Microsoft Research Asia, Beijing
Venue:
IEEE Transactions on Pattern Analysis and Machine Intelligence
Year:
2012

Citing 0
Cited 13

TILT: Transform Invariant Low-Rank Textures

International Journal of Computer Vision
Robust sparse bounding sphere for 3D face recognition

Image and Vision Computing
Lucas-Kanade based entropy congealing for joint face alignment

Image and Vision Computing
Robust Visual Tracking via Structured Multi-Task Sparse Learning

International Journal of Computer Vision
An ADM-based splitting method for separable convex programming

Computational Optimization and Applications
Efficient face recognition using wavelet-based generalized neural network

Signal Processing
Digital paparazzi: spotting celebrities in professional photo libraries

ACCV'12 Proceedings of the 11th Asian conference on Computer Vision - Volume Part II
Robust image annotation via simultaneous feature and sample outlier pursuit

ACM Transactions on Multimedia Computing, Communications, and Applications (TOMCCAP)
Motion matters: a novel framework for compressing surveillance videos

Proceedings of the 21st ACM international conference on Multimedia
Low-rank representation with local constraint for graph construction

Neurocomputing
A scalable approach to column-based low-rank matrix approximation

IJCAI'13 Proceedings of the Twenty-Third international joint conference on Artificial Intelligence
Face recognition for web-scale datasets

Computer Vision and Image Understanding
Robust subspace discovery via relaxed rank minimization

Neural Computation

Quantified Score

Hi-index	0.14

Visualization

Abstract

This paper studies the problem of simultaneously aligning a batch of linearly correlated images despite gross corruption (such as occlusion). Our method seeks an optimal set of image domain transformations such that the matrix of transformed images can be decomposed as the sum of a sparse matrix of errors and a low-rank matrix of recovered aligned images. We reduce this extremely challenging optimization problem to a sequence of convex programs that minimize the sum of \ell^1-norm and nuclear norm of the two component matrices, which can be efficiently solved by scalable convex optimization techniques. We verify the efficacy of the proposed robust alignment algorithm with extensive experiments on both controlled and uncontrolled real data, demonstrating higher accuracy and efficiency than existing methods over a wide range of realistic misalignments and corruptions.