TY - JOUR
T1 - Robust correlated and individual component analysis
AU - Panagakis, Yannis
AU - Nicolaou, Mihalis A.
AU - Zafeiriou, Stefanos
AU - Pantic, Maja
N1 - Indexed keywords
Engineering controlled terms: Behavioral research; Face recognition; Gaussian noise (electronic); Optimization
Canonical correlation analysis; Individual components; low-rank; sparsity; Time warping; Engineering main heading: Modal analysis
PY - 2016/8/1
Y1 - 2016/8/1
N2 - Recovering correlated and individual components of two, possibly temporally misaligned, sets of data is a fundamental task in disciplines such as image, vision, and behavior computing, with application to problems such as multi-modal fusion (via correlated components), predictive analysis, and clustering (via the individual ones). Here, we study the extraction of correlated and individual components under real-world conditions, namely i) the presence of gross non-Gaussian noise and ii) temporally misaligned data. In this light, we propose a method for the Robust Correlated and Individual Component Analysis (RCICA) of two sets of data in the presence of gross, sparse errors. We furthermore extend RCICA in order to handle temporal incongruities arising in the data. To this end, two suitable optimization problems are solved. The generality of the proposed methods is demonstrated by applying them onto 4 applications, namely i) heterogeneous face recognition, ii) multi-modal feature fusion for human behavior analysis (i.e., audio-visual prediction of interest and conflict), iii) face clustering, and iv) thetemporal alignment of facial expressions. Experimental results on 2 synthetic and 7 real world datasets indicate the robustness and effectiveness of the proposed methodson these application domains, outperforming other state-of-the-art methods in the field. © 1979-2012 IEEE.
AB - Recovering correlated and individual components of two, possibly temporally misaligned, sets of data is a fundamental task in disciplines such as image, vision, and behavior computing, with application to problems such as multi-modal fusion (via correlated components), predictive analysis, and clustering (via the individual ones). Here, we study the extraction of correlated and individual components under real-world conditions, namely i) the presence of gross non-Gaussian noise and ii) temporally misaligned data. In this light, we propose a method for the Robust Correlated and Individual Component Analysis (RCICA) of two sets of data in the presence of gross, sparse errors. We furthermore extend RCICA in order to handle temporal incongruities arising in the data. To this end, two suitable optimization problems are solved. The generality of the proposed methods is demonstrated by applying them onto 4 applications, namely i) heterogeneous face recognition, ii) multi-modal feature fusion for human behavior analysis (i.e., audio-visual prediction of interest and conflict), iii) face clustering, and iv) thetemporal alignment of facial expressions. Experimental results on 2 synthetic and 7 real world datasets indicate the robustness and effectiveness of the proposed methodson these application domains, outperforming other state-of-the-art methods in the field. © 1979-2012 IEEE.
KW - EWI-27126
KW - HMI-HF: Human Factors
KW - Sparsity
KW - canonical correlation analysis
KW - Low rank
KW - time warping
KW - EC Grant Agreement nr.: FP7/611153
KW - IR-103791
KW - Multi-modal analysis
KW - individual components
KW - n/a OA procedure
U2 - 10.1109/TPAMI.2015.2497700
DO - 10.1109/TPAMI.2015.2497700
M3 - Article
SN - 0162-8828
VL - 38
SP - 1665
EP - 1678
JO - IEEE transactions on pattern analysis and machine intelligence
JF - IEEE transactions on pattern analysis and machine intelligence
IS - 8
ER -