Abstract
Understanding a social scene requires not only perceiving individuals, but also detecting and understanding social interactions between them. Humans are adept at perceiving social interactions, an ability that develops early in infancy (Hamlin et al., 2007) and is shared with other primates (Sliwa and Freiwald 2017). We recently identified a region of the human posterior superior temporal sulcus (pSTS) that is selectively engaged when people view third-party social interactions (Isik et al., 2017). These findings underscore the importance of perceiving social interactions, but leave unanswered the question of how quickly and automatically it occurs. Is social interaction detection a rapid, feedforward perceptual process, or a slower post-perceptual inference? To answer this question, we used magnetoencephalography (MEG) decoding to ask when the human brain detects third-party social interactions. In particular, subjects in the MEG viewed snapshots of visually matched real-world scenes containing a pair of people who were either engaged in a social interaction or acting independently (Figure 1). To separate decoding from task demands, subjects performed an orthogonal task. We could read out the presence versus absence of a social interaction from subjects' MEG data extremely quickly, as early as 125 ms after stimulus onset (Figure 2A). This decoding latency is similar to previously reported decoding latencies of primarily feedforward visual processes, such as invariant object recognition (Isik et al., 2014). Importantly, this decoding does not seem to be based on low-level image properties: the images are not decodable based on pixel intensity or the output of a V1-like model, and the social interaction decoding we observed occurs considerably later than the decoding of low-level image identity observed in the same subjects (Figure 2B). These results suggest that the detection of social interactions is a rapid feedforward perceptual process, rather than a slow post-perceptual inference.
Meeting abstract presented at VSS 2018