The selective attention for identification model (SAIM): A framework for closing the gap between the behavioural and neurological levels

The selective attention for identification model (SAIM) implements translation-invariant object recognition in multiple-object scenes by employing competitive and cooperative interactions between rate code neurons. With these mechanisms, SAIM can model a wide range of experimental evidence on attention and its disorders (e.g., Heinke & Humphreys, 2003). This chapter presents two new versions of SAIM that address issues at the two sides of the gap between the behavioural and neurological levels. For the behavioural side, we demonstrate that an extension of SAIM can process ecologically valid inputs (natural colour images). For the neurological side, we present a method for replacing rate code neurons with spiking neurons while maintaining SAIM’s successes on the behavioural level. Hence, this method allows SAIM to link attentional behaviour with processes on the neurological level, and vice versa.