Gaze is a relevant cue in social context, in both aspects of general gaze behaviour and gaze interaction. The research line we are pursuing considers gaze behaviour analysis and simulation from the perspective of stochastic processes [1, 2] and optimal foraging theories [3]. In general, gaze shifts dynamics in attentional allocation is determined not just by stimulus-driven selection but, importantly, by social value as modulating the selection history of relevant multimodal items.
In ongoing research, we are addressing the deployment of perceptual attention to social interactions as displayed in conversational clips, when relying on multimodal information (audio and video) [6]. From a broader point of view, affective value should also be taken into account in the modulation of gaze behaviour [4].
Beyond contingent social contex, gaze behaviour can be exploited as a marker of long lasting individual traits. Machine learning techniques have been exploited, as opposed to classic statistical tools, to find a relation between gaze patterns and personality [5] in the effort to bridge the gap between available data and models. Typical approaches focus on the analysis of spatial and temporal preferences of gaze deployment over specific regions of the observed face adopting classic statistical methods. In [5] we proposed a different analysis perspective based on novel data-mining techniques and a probabilistic classification method that relies on Gaussian Processes exploiting Automatic Relevance Determination (ARD) kernel.
