Development and validation of a new dynamic body expression database

Poster Presentation: Sunday, May 18, 2025, 2:45 – 6:45 pm, Banyan Breezeway
Session: Face and Body Perception: Body

Lilly Kelemen1 (), J. Brendan Ritchie1, Chris Baker1, Shruti Japee1; 1National Institute of Mental Health

Facial expressions are a crucial, nonverbal means of social communication. While much research has focused on the role of facial expressions in social interactions, the role of body expressions has been less studied. Further, studies examining the role of body cues often rely on static images of actors posing various body expressions with their faces digitally obscured or cropped. To address this gap, we developed the Facial and Body Movement Database (FaBMoD), which comprises 352 videos of nine (5 male) racially diverse professional actors making a variety of body movements. The actors were instructed to portray six basic emotional movements and four neutral movements. To understand the relative role of face and body expressions in social communication, we manipulated the visibility of facial information in two ways. First, each movement was performed in two orientations––once facing the camera and once facing away from the camera such that only a limited portion of the face was visible. Second, each movement was performed with and without masking using a green screen head covering. To validate the stimulus set, 10 participants completed a computer task where they were shown each video and asked to identify the expression portrayed and rate each video on a scale of 1-9 for intensity, genuineness, and valence. Preliminary results indicate good overall correspondence for most videos between the intended expression and the expression perceived by participants. This was especially the case for videos depicting anger and happiness, while videos intended to depict surprise were sometimes perceived as fear. Neutral movements were classified as neutral in almost every case. This new database of body movements, along with the validation data, will be used to understand the role of body expressions in social communication. We also intend to share this database with other researchers world-wide.