Affective Gait Recognition And Baseline Evaluation From Real World Samples

Over the years a lot of research efforts have been put into recognizing human emotions from facial expressions. However, in many scenarios access to suitable face data is difficult, and therefore there is a need for methodology that can be used when people are observed from a distance. A potential modality for this is human gait. Early attempts to recognize human emotion from gait have been limited to acted data. Furthermore, in these approaches the data has been captured in controlled settings. This paper presents the first experiments for automated affective gait recognition using non acted real world samples. A database of 96 subjects affected by positive or negative feedback is collected and two baseline methods are used to recognize the affective state of a person. The baseline results are promising and encourage further study in this domain.