TY - GEN
T1 - Action unit reconstruction of occluded facial expression
AU - Wu, Chung Hsien
AU - Lin, Jen Chun
AU - Wei, Wen Li
N1 - Publisher Copyright:
© 2014 IEEE.
PY - 2014/11/12
Y1 - 2014/11/12
N2 - Facial occlusion is a critical issue that may dramatically degrade the performance on facial expression-based emotion recognition. In this study, the Error Weighted Cross-Correlation Model (EWCCM) is employed to predict the facial Action Unit (AU) under partial facial occlusion from non-occluded facial regions for facial geometric feature reconstruction. In EWCCM, a Gaussian Mixture Model (GMM)-based Cross-Correlation Model (CCM) is first adopted to construct the statistical dependency among features from paired facial components such as eyebrows-cheeks of the non-occluded regions for AU prediction of the occluded region. A Bayesian classifier weighting scheme is then used to enhance the AU prediction accuracy considering the contributions of the GMM-based CCMs. Based on the predicted AU, a regression fusion scheme is proposed to reconstruct the occluded facial geometric features. Experimental results show that the proposed approach yielded satisfactory results on the NCKU-FEPO database for facial AU reconstruction.
AB - Facial occlusion is a critical issue that may dramatically degrade the performance on facial expression-based emotion recognition. In this study, the Error Weighted Cross-Correlation Model (EWCCM) is employed to predict the facial Action Unit (AU) under partial facial occlusion from non-occluded facial regions for facial geometric feature reconstruction. In EWCCM, a Gaussian Mixture Model (GMM)-based Cross-Correlation Model (CCM) is first adopted to construct the statistical dependency among features from paired facial components such as eyebrows-cheeks of the non-occluded regions for AU prediction of the occluded region. A Bayesian classifier weighting scheme is then used to enhance the AU prediction accuracy considering the contributions of the GMM-based CCMs. Based on the predicted AU, a regression fusion scheme is proposed to reconstruct the occluded facial geometric features. Experimental results show that the proposed approach yielded satisfactory results on the NCKU-FEPO database for facial AU reconstruction.
UR - http://www.scopus.com/inward/record.url?scp=84916243284&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84916243284&partnerID=8YFLogxK
U2 - 10.1109/ICOT.2014.6956628
DO - 10.1109/ICOT.2014.6956628
M3 - Conference contribution
AN - SCOPUS:84916243284
T3 - IEEE International Conference on Orange Technologies, ICOT 2014
SP - 177
EP - 180
BT - IEEE International Conference on Orange Technologies, ICOT 2014
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2014 IEEE International Conference on Orange Technologies, ICOT 2014
Y2 - 20 September 2014 through 23 September 2014
ER -