Abstract
Feature integration is one of the fundamental steps in vision. To examine temporal aspects of feature fusion, we displayed two verniers with opposite offset direction in rapid succession at the same location. Only one fused vernier is perceived with its offset being a combination of the two verniers. We balanced the duration of the two verniers at a performance level of 50% for each observer individually. To probe the time course of feature fusion, we masked the two verniers by a light mask at several SOAs. Mask presentation before the onset of the verniers causes the second vernier to dominate the fused vernier, whereas the first vernier dominates when the mask is presented after the onset of the second vernier. Hence, we can manipulate the contribution of each vernier offset to the perceived, fused vernier. Surprisingly, the mask we used does not mask a single vernier when presented alone. Thus, the modulation of dominance by the mask must be caused by an interaction with the process of integration.