Neural correlates of experience with CCTV surveillance of naturalistic prosocial and antisocial interactions: a reverse correlation analysis

Abstract

1Previous research using reverse correlation to explore the relationship 2 between brain activity and presented image information found that Face Fusiform 3 Area (FFA) activity could be related to the appearance of faces during free viewing of 4 the Hollywood movie "The Good, the Bad, and the Ugly" (Hasson, et al, 2004). We 5 applied this approach to the naturalistic viewing of unedited footage of city-centre 6 closed-circuit television (CCTV) surveillance. Two 300 second videos were used, one 7 containing prosocial activities and the other antisocial activities. Brain activity 8 revealed through fMRI as well as eye movements were recorded while fifteen expert 9 CCTV operators with a minimum of 6 months experience of CCTV surveillance 10 alongside an age and gender matched control group of fifteen novice viewers were 11 scanned while watching the videos. Independent scans functionally localized FFA 12 and posterior Superior Temporal Sulcus (pSTS) activity using faces/houses and 13 intact/scrambled point-light biological motion displays respectively. Reverse 14 correlation revealed peaks in FFA and pSTS brain activity corresponding to the 15 expert and novice eye movements directed towards faces and biological motion 16 across both videos. In contrast, troughs in activation corresponded to camera-17 induced motion when a clear view of visual targets were temporarily not available. 18Our findings, validated by the eye movement data, indicate that the predicted 19 modulation of brain activity occurs as a result of salient features of faces and 20 biological motion embedded within the naturalistic stimuli. The examination of 21 expertise revealed that in both pSTS and FFA the novices had significantly more 22 activated timeframes than the experienced observers for the prosocial video. 23 However, no difference was found for the antisocial video. The modulation of brain 24 activity, as well as the effect of expertise gives a novel insight into the underlying 25 visual processes in an applied real-life task. 26

Patterns of Saliency and Semantic Features Distinguish Gaze of Expert and Novice Viewers of Surveillance Footage

“…Previous studies have compared CCTV operators to novices when performing the task of judging harmful intent from CCTV videos. Most studies have found group differences between CCTV operators and novices in brain activity and in the ability to recognize and predict harmful intention (Gillard et al, 2019; Grant & Williams, 2011; Howard et al, 2009; Petrini, McAleer, Neary, Gillard, & Pollick, 2014; but also see Troscianko, Holmes, Stillman, Mirmehdi, Wright, & Wilson, 2004). Hence, studying the differences in information processing between experienced CCTV operators and novices provides a unique window to unveil efficient strategies acquired by human experts through extensive learning.…”