WITHDRAWN: Facial expression recognition from videos using CNN and feature aggregation

Abstract

Abstract The facial expression recognition (FER) system has always been in trend and majorly when it comes to the FER from the video clips, then it becomes a crucial task to do. The differences seen in the visual descriptor and the emotions seen on the face need to be filled in the FER from videos. Here in our proposed work aggregation is done between the spatial and temporal convolutional features that are available in the whole video for the recognition of facial expressions in any video. We are using here 15–15 both spatial and temporal streams, in which every stream, i.e. spatial is corresponding with the temporal flow which creates a layer of aggregation for the training of end to end FER system from the video. This training gives a better representation of the video and can avoid the problem of over fitting from the limitation of available datasets. We have found that the proposed approach is best for the aggregation of the spatial–temporal features from video dataset as compared to other available methods. The dataset we have used are RML, MMI, BAUM-1 s, eNTERFACE05, FER-2013 and the results obtained after applying the proposed approach on this datasets are satisfactory.