2019/11/12 by Mariana-Iuliana Georgescu, Radu Tudor Ionescu, Georgescu, Mariana-Iuliana +1
Computer Science · Neuroscience · Psychology · #Computer Vision and Pattern Recognition (cs.CV) #Emotion and Mood Recognition #FOS: Computer and information sciences #Face Recognition and Perception #Face recognition and analysis #Machine Learning (cs.LG)
paper · pdf · doi:10.48550/arxiv.1911.04852
openalex publication_date 2019/11/12 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
In this paper, we present an approach based on convolutional neural networks\n(CNNs) for facial expression recognition in a difficult setting with severe\nocclusions. More specifically, our task is to recognize the facial expression\nof a person wearing a virtual reality (VR) headset which essentially occludes\nthe upper part of the face. In order to accurately train neural networks for\nthis setting, in which faces are severely occluded, we modify the training\nexamples by intentionally occluding the upper half of the face. This forces the\nneural networks to focus on the lower part of the face and to obtain better\naccuracy rates than models trained on the entire faces. Our empirical results\non two benchmark data sets, FER+ and AffectNet, show that our CNN models'\npredictions on lower-half faces are up to 13% higher than the baseline CNN\nmodels trained on entire faces, proving their suitability for the VR setting.\nFurthermore, our models' predictions on lower-half faces are no more than 10%\nunder the baseline models' predictions on full faces, proving that there are\nenough clues in the lower part of the face to accurately predict facial\nexpressions.\n