非表示:
キーワード:
-
要旨:
Local structures, presented briefly in the form of a scrambled image, prime recognition of a target picture (Tjan et al, 2000 ARVO). Does this priming reveal a mechanism intrinsic to rapid scene perception?
Participants had to judge which of two black-and-white cut-outs belong to a natural-scene picture embedded in a presentation sequence. Three types of images could be presented in combinations within the sequence: a coherent colour picture (C), its block-scrambled (46 × 46 pixels) version (J), or its pixel-scrambled colour-preserving version (F). Prior studies showed that performance for 3J2C (three frames of J followed by two frames of C, 14 ms frame-1) is better than for 2C alone.
In the present study, we found that performance decreased if either the 3J or the 2C component of 3J2C was removed by replacing it with 3F or 2F, respectively. Performance also worsened when observers were shown two presentation sequences, one containing only the 3J and the other only 2C, before making a decision. Taken together, these results show that the level of facilitation by local structures (3J) is beyond a simple combination of two sources of information (3J and 2C). Local structures prime scene recognition by interacting with the normal process for rapid scene perception.
