Cross-cultural differences in recognizing affect from body posture

Abstract

Conveyance and recognition of human emotion and affective expression is influenced by many factors, including culture. Within the user modeling field, it has become increasingly necessary to understand the role affect can play in personalizing interactive interfaces using embodied animated agents. However, little research within the computer science field aims at understanding cultural differences within this vein. Therefore, we conducted a study to evaluate if differences exist in the way various cultures perceive emotion from body posture. We used static posture images of affectively expressive avatars to conduct recognition experiments with subjects from three cultures. After analyzing the subjects' judgments using multivariate analysis, we grounded the identified differences into a set of low-level posture features. We then used Mixture Discriminant Analysis (MDA) and an unsupervised expectation maximization (EM) model to build separate cultural models for affective posture recognition. Our results could prove useful to aide designers in creating more effective affective avatars.

Multi-score Learning for Affect Recognition: The Case of Body Postures

“…This vector describes the configuration of the posture in terms of distances between body joints and angles between body segments. Details on the data collection are provided in [10]. The second dataset contains 103 non-acted postures collected in a whole-body computer game scenario.…”

Automatic Recognition of Affective Body Movement in a Video Game Scenario

“…Until recently, the main modality used to measure the affective state of people was their facial expressions [4]. Recent psychology studies, however, have revealed that body expressions are also a very good indicator of affect [e.g., 2,3,10]. These studies encouraged us into researching the possibility to create an automatic recognition system that would use the players' body movement to detect their affective state.…”

“…Using acted affective postures, Kleinsmith et al [24] explored cultural differences in expressing and recognizing affect from body expressions. The analysis, based on the set of low-level descriptive features proposed in [24], highlighted some differences between the cultures but showed also the possibility to build automatic recognition models that reflect the recognition of human observers from different cultures.…”

“…The analysis, based on the set of low-level descriptive features proposed in [24], highlighted some differences between the cultures but showed also the possibility to build automatic recognition models that reflect the recognition of human observers from different cultures. Similar results were obtained for the Japanese culture on affective dimensions as discussed in [25].…”

“…In all theses studies, like many others [6,9,10,11,7], the affective states are acted and hence very stereotypical and even exaggerated making the generalization of these studies to real application scenarios more difficult.…”