Emotional Analysis on Facial Expression in the Wild (2014 - 2017)



This project will investigate an automatic machine recognition method for facial expression analysis in-the-wild. The facial image under unconstrained conditions will be normalized and reconstructed to a canonical face under a controlled environment where 1) illumination is uniform, 2) head-pose is in frontal view and 3) there is no occlusions. Meanwhile, the normalized faces should remain the facial expression cues. The project is funded by University of Portsmouth and the Engineering and Physical Sciences Research Council (EPSRC) .


Wang, Yiming, Hui Yu, Junyu Dong, Muwei Jian, and Honghai Liu. “Cascade Support Vector Regression-based Facial Expresssion-Aware Face Frontalization.” Accepted by ICIP (2017).
Paper (pdf)

Wang, Yiming, Hui Yu, Junyu Dong, Brett Stevens, and Honghai Liu. “Facial expression-aware face frontalization.” In Asian Conference on Computer Vision, pp. 375-388. Springer, Cham, 2016.
Paper (pdf)


Visual Computing Group @ University of Portsmouth