Static facial expression analysis in tough conditions: Data, evaluation protocol and benchmark

Abstract

Quality data recorded in varied realistic environments is vital for effective human face related research. Currently available datasets for human facial expression analysis have been generated in highly controlled lab environments. We present a new static facial expression database Static Facial Expressions in the Wild (SFEW) extracted from a temporal facial expressions database Acted Facial Expressions in the Wild (AFEW) [9], which we have extracted from movies. In the past, many robust methods have been reported in the literature. However, these methods have been experimented on different databases or using different protocols within the same databases. The lack of a standard protocol makes it difficult to compare systems and acts as a hindrance in the progress of the field. Therefore, we propose a person independent training and testing protocol for expression recognition as part of the BEFIT workshop. Further, we compare our dataset with the JAFFE and Multi-PIE datasets and provide baseline results.

Riesz-based Volume Local Binary Pattern and A Novel Group Expression Model for Group Happiness Intensity Analysis

“…Since the HAPPEI database [7] is the only database for group happiness intensity analysis, we apply the following strategy for analyzing our proposed methods. Firstly, we examine the performance of the RVLBP descriptor on three challenging face databases [5,7,31] in Section 4.1. Secondly, we conduct the experiment on the HAPPEI database [7] for evaluating GEM CCRF in Section 4.2.…”

“…Recently, continuous conditional random fields has been proposed to The blue circle on a facial image represents the extracted content (including local and global attributes), z i is an happiness intensity label, an edge (a solid line) between z i and z j , e.g. g 2,5 , means the dependency between intensities of two faces, an edge (a dash line), e.g. f 5 , represents the dependency of an intensity label on its content.…”

“…In order to evaluate the performance of the RVLBP descriptor, we conduct experiments on the SFEW [5], GENKI-4K [31] and HAPPEI [7] databases, for the task of facial expression recognition in the wild, smile detection and happiness intensity estimation, respectively.…”

“…(1) Based on [5], the experiments on the SFEW database follow the strictly person independent strategy protocol (i.e. the train and test set have no common subjects).…”

Emotion recognition in the wild

“…EmotiW [1] added more data to the AFEW database, including more complex data to further stimulate research in affective computing towards real-world conditions. During EmotiW 2015 [5], a new sub-challenge-image based static facial expression recognition-was introduced, which was based on the Static Facial Expressions in the Wild (SFEW) database [3]. The SFEW database has been extracted from the AFEW database using a fiducial points based clustering technique.…”