Human Perception of Laughter from Context-Free Whole Body Motion Dynamic Stimuli

Perception and Automatic Recognition of Laughter from Whole-Body Motion: Continuous and Categorical Perspectives

“…From Fig. 5 and Table 1 we can see that the first component strongly mirrors the findings of McKeown et al [50] and most likely reflects an intensity dimension ranging from non-laughter (zero intensity) to hilarious laughter (high intensity). Interpretation of further components is necessarily speculative since only perceived intensity was recorded by McKeown et al…”

“…To this end, laughs were selected that spanned a wide range of experimenterjudged triggers, contexts and intensities. For an analysis of the relationship between observer judgements and experimenter judgements, which were made using such privileged contextual and multimodal information, see [50]. These stimuli (UCL Body Motion Stimuli) are also available at the ILHAIRE laughter database: www.qub.ac.uk/ ilhairelaughter [51].…”

“…It may be that the categorisations reported by observers were the result of their fitting our, somewhat pragmatic, range of categories onto their fundamental perception of the laughter. Evidence for this hypothesis comes from a study by McKeown et al using this same stimulus set [50]. In their paper, in addition to categorizing the stimuli according to the same five categories, observers also rated the intensity of the laughter in each animation.…”

Laughter Type Recognition from Whole Body Motion

“…The sex, age, cultural background and personality of the laughter and observer should also be further considered, for example, laughter produced by extroverts and introverts may vary and specific attitudes towards laughter may affect the perception of the emotional content of the laughter. Some of these factors have been investigated in [32] using the same set of body laughter stimuli used in our study. The role of body movements may be more complex in multimodal displays than in this uni-modal study and our findings should be validated with simultaneous facial and audio information to establish their applicability in functional human-avatar interactions.…”

Laughter Research: A Review of the ILHAIRE Project

“…Two data gathering sessions were dedicated to collecting this data. One in Belfast developed the laughter induction techniques (McKeown et al [72]), and was followed by a similar session conducted at UCL in which the motion capture data was made available as part of the database (Griffin et al [40]; McKeown et al [71]). The available data consists of 126 animated "stick figure" video stimuli of laughter that has been categorized as either hilarious, social, fake, awkward or not a laugh.…”