Abstract
Purpose: In spatiotemporal boundary formation (SBF), surface boundaries, object shape, and global motion are perceived from sequences of local element changes (Shipley & Kellman, 1994a, 1994b). It has previously been suggested that SBF uses sequential changes as motion signals, and these feed into a computation that determines edges and eventually form (Shipley & Kellman, 1996). A model built on these assumptions is sufficient to model human performance in an edge discrimination test where the edges are defined by SBF (Erlikhman et al., 2012). However, no previous work has examined the neural correlates of SBF. Method: In a series of experiments, we applied EEG techniques to examine the neural timecourse of the formation of a shape representation in SBF displays. In one condition, four SBF shapes defined by color changes of small dot elements traveled on a circular path. In a control condition, the same number of changes occurred on every frame, but the sequence of the changes was randomized and no SBF shapes are perceived. We collected EEG data from 10 subjects. Results: The difference waveform elicited by subtracting the event related response to the control stimuli from that to the SBF stimuli consisted of a significant negative component from 200-300ms followed by a significant positive component from ~300-400ms. Control experiments were performed to test the hypothesis that the earlier component reflects the extraction of the motion signals necessary for SBF (e.g., Kuba & Kubova, 1992) and the later component reflects the subsequent formation of the perceived shapes produced by SBF.
Meeting abstract presented at VSS 2013