A distinction can be drawn between models of crowding that rely on information loss, with crowded items either suppressed [ 11, 12] or lost [ 13 15], and change-based models such as averaging [ 16] and flanker substitution [ 20 22].
Using computational models of crowds, it is possible to set up various scenarios and evaluate how the occupants move through the building.The trouble is that the software that is used to do this generally treats people like particles in a fluid, says Mr Wittasek.
Pooling and nonpooling, together, make up the space of all possible models of crowding.
These changes follow predictions from two change-based models of crowding.
Pooling and nonpooling (simple substitution) together include all possible models of crowding.
Our results are also inconsistent with inhibition-based models of crowding [ 11, 12] and explanations based on masking [ 17, 29 31].
These results confirm predictions from change-based models of crowding, such as averaging [ 16], and establish crowding as a regularization process that simplifies the peripheral field by promoting consistent appearance among adjacent objects.
For example, integration and averaging models of crowding (Parkes et al., 2001; Greenwood et al., 2009) propose that, while isolated contours are processed by simple cells, a high concentration of flanking contours in a small region or 'integration field' (Pelli et al., 2004) leads to a greater response of complex cells which then suppress simple cell activity within their receptive field area.
A force that generally follows the standard Mediterranean model of crowd control ("keep out of the way and then, when something happens, respond with overwhelming force," as David Swift, of the Association of Chief Police Officers, puts it) will test a softer line.
A basic model of crowd movement along evacuation paths is developed and allows detailed analysis of the effect of the initial crowd density on congestion.
